From 34869e27aef41cc07dce4a246604799adf3d02b8 Mon Sep 17 00:00:00 2001
From: Nathan Willis <nwillis@glyphography.com>
Date: Sun, 26 Feb 2023 15:12:02 +0000
Subject: [PATCH] Wrap all acronyms of 3+ letters in abbr tags for CSS smallcap
 styling, exlcuding tt, code, and literal quotations.

---
 README.md                                     |  4 +-
 character-tables/character-tables-arabic.md   | 10 ++--
 character-tables/character-tables-bengali.md  |  2 +-
 .../character-tables-devanagari.md            |  2 +-
 character-tables/character-tables-gujarati.md |  2 +-
 character-tables/character-tables-gurmukhi.md |  2 +-
 character-tables/character-tables-kannada.md  |  2 +-
 character-tables/character-tables-khmer.md    | 13 ++---
 .../character-tables-malayalam.md             | 12 ++---
 .../character-tables-mongolian.md             |  4 +-
 character-tables/character-tables-myanmar.md  | 12 ++---
 character-tables/character-tables-nko.md      |  6 +--
 character-tables/character-tables-oriya.md    |  2 +-
 character-tables/character-tables-sinhala.md  |  2 +-
 character-tables/character-tables-syriac.md   | 10 ++--
 character-tables/character-tables-tamil.md    |  2 +-
 character-tables/character-tables-telugu.md   |  2 +-
 character-tables/character-tables-tibetan.md  | 12 ++---
 errata.md                                     | 20 ++++----
 opentype-shaping-arabic-general.md            | 28 +++++------
 opentype-shaping-arabic.md                    | 34 ++++++-------
 opentype-shaping-bengali.md                   | 14 +++---
 opentype-shaping-default.md                   | 12 ++---
 opentype-shaping-devanagari.md                | 14 +++---
 opentype-shaping-emoji.md                     | 42 ++++++++--------
 opentype-shaping-gujarati.md                  | 14 +++---
 opentype-shaping-gurmukhi.md                  | 14 +++---
 opentype-shaping-hangul.md                    |  4 +-
 opentype-shaping-hebrew.md                    | 18 +++----
 opentype-shaping-indic-general.md             | 16 +++---
 opentype-shaping-kannada.md                   | 12 ++---
 opentype-shaping-khmer.md                     | 25 +++++-----
 opentype-shaping-malayalam.md                 | 14 +++---
 opentype-shaping-mongolian.md                 | 20 ++++----
 opentype-shaping-myanmar.md                   | 25 +++++-----
 opentype-shaping-nko.md                       | 30 +++++------
 opentype-shaping-normalization.md             |  2 +-
 opentype-shaping-oriya.md                     | 14 +++---
 opentype-shaping-sinhala.md                   | 14 +++---
 opentype-shaping-syriac.md                    | 28 +++++------
 opentype-shaping-tamil.md                     | 14 +++---
 opentype-shaping-telugu.md                    | 14 +++---
 opentype-shaping-thai-lao.md                  | 50 +++++++++----------
 opentype-shaping-tibetan.md                   | 12 ++---
 opentype-shaping-use.md                       | 40 +++++++--------
 45 files changed, 324 insertions(+), 321 deletions(-)

diff --git a/README.md b/README.md
index 633cd52..c2f029b 100644
--- a/README.md
+++ b/README.md
@@ -17,7 +17,7 @@ _<aside>Thanks also to the developers of HarfBuzz and AllSorts, plus many other
 
 At present, we are seeking comments and bugfixes on the Indic-script,
 Arabic-like, Hangul, Hebrew, Thai/Lao, Tibetan, Khmer, Myanmar,
-default, and USE documents. Interested readers and contributors can
+default, and <abbr>USE</abbr> documents. Interested readers and contributors can
 begin at the
 
   - [Indic General](opentype-shaping-indic-general.md) 
@@ -31,7 +31,7 @@ begin at the
   - [Thai and Lao](opentype-shaping-thai-lao.md)
   - [Tibetan](opentype-shaping-tibetan.md)
   - [Myanmar](opentype-shaping-myanmar.md)
-  - [Universal Shaping Engine (USE)](opentype-shaping-use.md)
+  - [Universal Shaping Engine (<abbr>USE</abbr>)](opentype-shaping-use.md)
     - All complex scripts that are not handled by a dedicated
       script-specific shaping model
   - [Default](opentype-shaping-default.md)
diff --git a/character-tables/character-tables-arabic.md b/character-tables/character-tables-arabic.md
index d98fde0..51a7515 100644
--- a/character-tables/character-tables-arabic.md
+++ b/character-tables/character-tables-arabic.md
@@ -41,8 +41,8 @@ that sequences of adjacent marks can be reordered as required by the
 orthography. 
 
 For Arabic, a subset of marks in the 220 and 230 classes are also
-designated _Modifier Combining Marks_ (MCM). These are denoted with
-_220_MCM_ and _230_MCM_ in the _Mark class_ column. The MCM marks are
+designated _Modifier Combining Marks_ (<abbr>MCM</abbr>). These are denoted with
+_220_MCM_ and _230_MCM_ in the _Mark class_ column. The <abbr>MCM</abbr> marks are
 treated differently during the mark-reordering stage.
 
 
@@ -643,7 +643,7 @@ fashion; shaping engines should cope with this situation gracefully.
 
 
 
-The combining grapheme joiner (CGJ) is primarily used to alter the
+The combining grapheme joiner (<abbr>CGJ</abbr>) is primarily used to alter the
 order in which adjacent marks are positioned during the
 mark-reordering stage, in order to adhere to the needs of a
 non-default language orthography.
@@ -661,12 +661,12 @@ sequence "ZWJ,_Letter_,ZWJ" would be used.
 
 <!--- Zero-Width Non Joiner explanation --->
 
-The right-to-left mark (RLM) and left-to-right mark (LRM) are used by
+The right-to-left mark (<abbr>RLM</abbr>) and left-to-right mark (<abbr>LRM</abbr>) are used by
 the Unicode bidirectionality algorithm (BiDi) to indicate the points
 in a text run at which the writing direction changes.
 
 
-<!--- How shaping is affected by the LTR and RTL markers explanation --->
+<!--- How shaping is affected by the <abbr>LTR</abbr> and <abbr>RTL</abbr> markers explanation --->
 
 
 The no-break space is primarily used to display those codepoints that
diff --git a/character-tables/character-tables-bengali.md b/character-tables/character-tables-bengali.md
index fbcfebc..ece1cf7 100644
--- a/character-tables/character-tables-bengali.md
+++ b/character-tables/character-tables-bengali.md
@@ -298,7 +298,7 @@ A secondary usage of the zero-width joiner is to prevent the formation of
 where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
diff --git a/character-tables/character-tables-devanagari.md b/character-tables/character-tables-devanagari.md
index bbfccd5..a4c2e82 100644
--- a/character-tables/character-tables-devanagari.md
+++ b/character-tables/character-tables-devanagari.md
@@ -331,7 +331,7 @@ A secondary usage of the zero-width joiner is to prevent the formation of
 where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
diff --git a/character-tables/character-tables-gujarati.md b/character-tables/character-tables-gujarati.md
index 7dac6bb..67b667e 100644
--- a/character-tables/character-tables-gujarati.md
+++ b/character-tables/character-tables-gujarati.md
@@ -299,7 +299,7 @@ A secondary usage of the zero-width joiner is to prevent the formation of
 where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
diff --git a/character-tables/character-tables-gurmukhi.md b/character-tables/character-tables-gurmukhi.md
index 46d88e2..3392237 100644
--- a/character-tables/character-tables-gurmukhi.md
+++ b/character-tables/character-tables-gurmukhi.md
@@ -299,7 +299,7 @@ A secondary usage of the zero-width joiner is to prevent the formation of
 where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
diff --git a/character-tables/character-tables-kannada.md b/character-tables/character-tables-kannada.md
index cf3a3cf..1a9457c 100644
--- a/character-tables/character-tables-kannada.md
+++ b/character-tables/character-tables-kannada.md
@@ -299,7 +299,7 @@ A secondary usage of the zero-width joiner is to prevent the formation of
 where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
diff --git a/character-tables/character-tables-khmer.md b/character-tables/character-tables-khmer.md
index 25abd8b..9633c05 100644
--- a/character-tables/character-tables-khmer.md
+++ b/character-tables/character-tables-khmer.md
@@ -262,12 +262,13 @@ A secondary usage of the zero-width joiner is to prevent the formation of
 where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 
-The no-break space is primarily used to display those codepoints that
-are defined as non-spacing (marks, dependent vowels (matras),
-below-base consonant forms, and post-base consonant forms) in an
-isolated context, as an alternative to displaying them superimposed on
-the dotted-circle placeholder. These sequences will match
-"NBSP,ZWJ,Halant,_Consonant_", "NBSP,_mark_", or "NBSP,_matra_".
+The no-break space (<abbr>NBSP<.abbr>) is primarily used to display
+those codepoints that are defined as non-spacing (marks, dependent
+vowels (matras), below-base consonant forms, and post-base consonant
+forms) in an isolated context, as an alternative to displaying them
+superimposed on the dotted-circle placeholder. These sequences will
+match "NBSP,ZWJ,Halant,_Consonant_", "NBSP,_mark_", or
+"NBSP,_matra_".
 
 In addition to general punctuation, runs of Khmer text often use the
 danda (`U+0964`) and double danda (`U+0965`) punctuation marks from
diff --git a/character-tables/character-tables-malayalam.md b/character-tables/character-tables-malayalam.md
index d427c63..01b78e0 100644
--- a/character-tables/character-tables-malayalam.md
+++ b/character-tables/character-tables-malayalam.md
@@ -299,10 +299,10 @@ A secondary usage of the zero-width joiner is to prevent the formation of
 where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 
-The no-break space is primarily used to display those codepoints that
-are defined as non-spacing (marks, dependent vowels (matras),
-below-base consonant forms, and post-base consonant forms) in an
-isolated context, as an alternative to displaying them superimposed on
-the dotted-circle placeholder. These sequences will match
-"NBSP,ZWJ,Halant,_Consonant_", "NBSP,_mark_", or "NBSP,_matra_".
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display
+those codepoints that are defined as non-spacing (marks, dependent
+vowels (matras), below-base consonant forms, and post-base consonant
+forms) in an isolated context, as an alternative to displaying them
+superimposed on the dotted-circle placeholder. These sequences will
+match "NBSP,ZWJ,Halant,_Consonant_", "NBSP,_mark_", or "NBSP,_matra_".
 
diff --git a/character-tables/character-tables-mongolian.md b/character-tables/character-tables-mongolian.md
index a8bb011..9e8d058 100644
--- a/character-tables/character-tables-mongolian.md
+++ b/character-tables/character-tables-mongolian.md
@@ -38,8 +38,8 @@ that sequences of adjacent marks can be reordered as required by the
 orthography. 
 
 For Mongolian, a subset of marks in the 220 and 230 classes are also
-designated _Modifier Combining Marks_ (MCM). These are denoted with
-_220_MCM_ and _230_MCM_ in the _Mark class_ column. The MCM marks are
+designated _Modifier Combining Marks_ (<abbr>MCM</abbr>). These are denoted with
+_220_MCM_ and _230_MCM_ in the _Mark class_ column. The <abbr>MCM</abbr> marks are
 treated differently during the mark-reordering stage.
 
 
diff --git a/character-tables/character-tables-myanmar.md b/character-tables/character-tables-myanmar.md
index 7659951..78a70ad 100644
--- a/character-tables/character-tables-myanmar.md
+++ b/character-tables/character-tables-myanmar.md
@@ -399,10 +399,10 @@ A secondary usage of the zero-width joiner is to prevent the formation of
 where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 
-The no-break space is primarily used to display those codepoints that
-are defined as non-spacing (marks, dependent vowels (matras),
-below-base consonant forms, and post-base consonant forms) in an
-isolated context, as an alternative to displaying them superimposed on
-the dotted-circle placeholder. These sequences will match
-"NBSP,ZWJ,Halant,_Consonant_", "NBSP,_mark_", or "NBSP,_matra_".
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display
+those codepoints that are defined as non-spacing (marks, dependent
+vowels (matras), below-base consonant forms, and post-base consonant
+forms) in an isolated context, as an alternative to displaying them
+superimposed on the dotted-circle placeholder. These sequences will
+match "NBSP,ZWJ,Halant,_Consonant_", "NBSP,_mark_", or "NBSP,_matra_".
 
diff --git a/character-tables/character-tables-nko.md b/character-tables/character-tables-nko.md
index d658937..e5caa8e 100644
--- a/character-tables/character-tables-nko.md
+++ b/character-tables/character-tables-nko.md
@@ -144,7 +144,7 @@ fashion; shaping engines should cope with this situation gracefully.
 
 
 
-The combining grapheme joiner (CGJ) is primarily used to alter the
+The combining grapheme joiner (<abbr>CGJ</abbr>) is primarily used to alter the
 order in which adjacent marks are positioned during the
 mark-reordering stage, in order to adhere to the needs of a
 non-default language orthography.
@@ -162,12 +162,12 @@ sequence "ZWJ,_Letter_,ZWJ" would be used.
 
 <!--- Zero-Width Non Joiner explanation --->
 
-The right-to-left mark (RLM) and left-to-right mark (LRM) are used by
+The right-to-left mark (<abbr>RLM</abbr>) and left-to-right mark (<abbr>LRM</abbr>) are used by
 the Unicode bidirectionality algorithm (BiDi) to indicate the points
 in a text run at which the writing direction changes.
 
 
-<!--- How shaping is affected by the LTR and RTL markers explanation --->
+<!--- How shaping is affected by the <abbr>LTR</abbr> and <abbr>RTL</abbr> markers explanation --->
 
 
 The no-break space is primarily used to display those codepoints that
diff --git a/character-tables/character-tables-oriya.md b/character-tables/character-tables-oriya.md
index 5cd9631..478a033 100644
--- a/character-tables/character-tables-oriya.md
+++ b/character-tables/character-tables-oriya.md
@@ -299,7 +299,7 @@ A secondary usage of the zero-width joiner is to prevent the formation of
 where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
diff --git a/character-tables/character-tables-sinhala.md b/character-tables/character-tables-sinhala.md
index eaf9f15..0739249 100644
--- a/character-tables/character-tables-sinhala.md
+++ b/character-tables/character-tables-sinhala.md
@@ -323,7 +323,7 @@ Sinhala text. The <abbr>ZWNJ</abbr> is referenced below in various regular
 expressions and shaping rules, however, because it is used by other
 Indic scripts.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
diff --git a/character-tables/character-tables-syriac.md b/character-tables/character-tables-syriac.md
index 0989641..d71bac7 100644
--- a/character-tables/character-tables-syriac.md
+++ b/character-tables/character-tables-syriac.md
@@ -38,8 +38,8 @@ that sequences of adjacent marks can be reordered as required by the
 orthography. 
 
 For Syriac, a subset of marks in the 220 and 230 classes are also
-designated _Modifier Combining Marks_ (MCM). These are denoted with
-_220_MCM_ and _230_MCM_ in the _Mark class_ column. The MCM marks are
+designated _Modifier Combining Marks_ (<abbr>MCM</abbr>). These are denoted with
+_220_MCM_ and _230_MCM_ in the _Mark class_ column. The <abbr>MCM</abbr> marks are
 treated differently during the mark-reordering stage.
 
 
@@ -197,7 +197,7 @@ does not encode a separate kashida character.
 | | | | | | |
 
 
-The combining grapheme joiner (CGJ) is primarily used to alter the
+The combining grapheme joiner (<abbr>CGJ</abbr>) is primarily used to alter the
 order in which adjacent marks are positioned during the
 mark-reordering stage, in order to adhere to the needs of a
 non-default language orthography.
@@ -215,12 +215,12 @@ sequence "ZWJ,_Letter_,ZWJ" would be used.
 
 <!--- Zero-Width Non Joiner explanation --->
 
-The right-to-left mark (RLM) and left-to-right mark (LRM) are used by
+The right-to-left mark (<abbr>RLM</abbr>) and left-to-right mark (<abbr>LRM</abbr>) are used by
 the Unicode bidirectionality algorithm (BiDi) to indicate the points
 in a text run at which the writing direction changes.
 
 
-<!--- How shaping is affected by the LTR and RTL markers explanation --->
+<!--- How shaping is affected by the <abbr>LTR</abbr> and <abbr>RTL</abbr> markers explanation --->
 
 
 The no-break space is primarily used to display those codepoints that
diff --git a/character-tables/character-tables-tamil.md b/character-tables/character-tables-tamil.md
index ecbf32b..82dff7d 100644
--- a/character-tables/character-tables-tamil.md
+++ b/character-tables/character-tables-tamil.md
@@ -399,7 +399,7 @@ A secondary usage of the zero-width joiner is to prevent the formation of
 where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
diff --git a/character-tables/character-tables-telugu.md b/character-tables/character-tables-telugu.md
index c2a5a58..5e341ba 100644
--- a/character-tables/character-tables-telugu.md
+++ b/character-tables/character-tables-telugu.md
@@ -299,7 +299,7 @@ A secondary usage of the zero-width joiner is to prevent the formation of
 where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
diff --git a/character-tables/character-tables-tibetan.md b/character-tables/character-tables-tibetan.md
index 52c8b13..6230399 100644
--- a/character-tables/character-tables-tibetan.md
+++ b/character-tables/character-tables-tibetan.md
@@ -318,9 +318,9 @@ A secondary usage of the zero-width joiner is to prevent the formation of
 where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 
-The no-break space is primarily used to display those codepoints that
-are defined as non-spacing (marks, dependent vowels (matras),
-below-base consonant forms, and post-base consonant forms) in an
-isolated context, as an alternative to displaying them superimposed on
-the dotted-circle placeholder. These sequences will match
-"NBSP,ZWJ,Halant,_consonant_", "NBSP,_mark_", or "NBSP,_matra_".
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display
+those codepoints that are defined as non-spacing (marks, dependent
+vowels (matras), below-base consonant forms, and post-base consonant
+forms) in an isolated context, as an alternative to displaying them
+superimposed on the dotted-circle placeholder. These sequences will
+match "NBSP,ZWJ,Halant,_consonant_", "NBSP,_mark_", or "NBSP,_matra_".
diff --git a/errata.md b/errata.md
index 914590d..3a15703 100644
--- a/errata.md
+++ b/errata.md
@@ -8,15 +8,15 @@ specification documents.
 **Table of Contents**
 
   - [Unicode](#unicode)
-      - [ZWJ and ZWNJ](#zwj-and-zwnj)
-	      - [Scope of ZWJ and ZWNJ](#scope-of-zwj-and-zwnj)
-	      - [ZWJ in redundant ligature lookups](#zwj-in-redundant-ligature-lookups)
+      - [<abbr>ZWJ</abbr> and <abbr>ZWNJ</abbr>](#zwj-and-zwnj)
+	      - [Scope of <abbr>ZWJ</abbr> and <abbr>ZWNJ</abbr>](#scope-of-zwj-and-zwnj)
+	      - [<abbr>ZWJ</abbr> in redundant ligature lookups](#zwj-in-redundant-ligature-lookups)
       - [Emoji](#emoji)
 	      - [Skin-tone permutations](#skin-tone-permutations)
 		  - [Gender permutations](#gender-permutations)
   - [OpenType](#opentype)
-      - [Null offsets in GSUB and GPOS](#null-offsets-in-gsub-and-gpos)
-      - [Sorting of GSUB and GPOS lookups](#sorting-of-gsub-and-gpos-lookups)
+      - [Null offsets in <abbr>GSUB</abbr> and <abbr>GPOS</abbr>](#null-offsets-in-gsub-and-gpos)
+      - [Sorting of <abbr>GSUB</abbr> and <abbr>GPOS</abbr> lookups](#sorting-of-gsub-and-gpos-lookups)
 	  - [Per-script applicability of feature tags](#per-script-applicability-of-feature-tags)
       - [Ordering of post-base and below-base consonants in Indic2 base-consonant determination](#ordering-of-post-base-and-below-base-consonants-in-indic2-base-consonant-determination)
       - [Lookup behavior](#lookup-behavior)
@@ -32,9 +32,9 @@ specification documents.
 
 This section lists errata pertaining to the Unicode Standard.
 
-### ZWJ and ZWNJ ###
+### <abbr>ZWJ</abbr> and <abbr>ZWNJ</abbr> ###
 
-#### Scope of ZWJ and ZWNJ ####
+#### Scope of <abbr>ZWJ</abbr> and <abbr>ZWNJ</abbr> ####
 
 Unicode provides the Zero Width Joiner (<abbr>ZWJ</abbr>) and Zero Width Non-Joiner
 (<abbr>ZWNJ</abbr>) control characters so that a text sequence can "request a
@@ -58,7 +58,7 @@ in the same "a,b,ZWNJ,c,d" sequence should inhibit the application of
 the ligature between "c" and "d".
 
 
-#### ZWJ in redundant ligature lookups ####
+#### <abbr>ZWJ</abbr> in redundant ligature lookups ####
 
 An "Implementation Notes" section in chapter 23.2 of the Unicode
 Standard says that font vendors should add <abbr>ZWJ</abbr> sequences to ligature
@@ -113,7 +113,7 @@ sequence "Woman,ZWJ,Man,ZWJ,Girl".
 
 This section lists errata pertaining to the OpenType specification.
 
-### Null offsets in GSUB and GPOS ###
+### Null offsets in <abbr>GSUB</abbr> and <abbr>GPOS</abbr> ###
 
 The headers of the <abbr>GSUB</abbr> and <abbr>GPOS</abbr> tables include fields that contain
 the offsets at which other structures within the font binary are
@@ -135,7 +135,7 @@ Furthermore, font-validation functions may overwrite a `NULL` into an
 offset field if the original value encountered was invalid.
 
 
-### Sorting of GSUB and GPOS lookups ###
+### Sorting of <abbr>GSUB</abbr> and <abbr>GPOS</abbr> lookups ###
 
 The OpenType specification requires that lookups in the <abbr>GSUB</abbr> table
 must be sorted into numeric order before they are applied.
diff --git a/opentype-shaping-arabic-general.md b/opentype-shaping-arabic-general.md
index 1d161a7..9bd41dc 100644
--- a/opentype-shaping-arabic-general.md
+++ b/opentype-shaping-arabic-general.md
@@ -17,9 +17,9 @@ Syriac, and Mongolian.
       - [2. Compound character composition and decomposition](#2-compound-character-composition-and-decomposition)
       - [3. Computing letter joining states](#3-computing-letter-joining-states)
       - [4. Applying the `stch` feature](#4-applying-the-stch-feature)
-      - [5. Applying the language-form substitution features from GSUB](#5-applying-the-language-form-substitution-features-from-gsub)
-      - [6. Applying the typographic-form substitution features from GSUB](#6-applying-the-typographic-form-substitution-features-from-gsub)
-      - [7. Applying the positioning features from GPOS](#7-applying-the-positioning-features-from-gpos)
+      - [5. Applying the language-form substitution features from <abbr>GSUB</abbr>](#5-applying-the-language-form-substitution-features-from-gsub)
+      - [6. Applying the typographic-form substitution features from <abbr>GSUB</abbr>](#6-applying-the-typographic-form-substitution-features-from-gsub)
+      - [7. Applying the positioning features from <abbr>GPOS</abbr>](#7-applying-the-positioning-features-from-gpos)
   
 
 
@@ -157,18 +157,18 @@ reordering the sequence into increasing numerical order.
 
 In addition, some Arabic and Syriac marks require special handling
 when shaping Arabic text, during the mark-reordering stage. These
-marks fall into two classes of _Modifier Combining Marks_ (MCM) that
+marks fall into two classes of _Modifier Combining Marks_ (<abbr>MCM</abbr>) that
 may need to be repositioned closer to the base character, when they
 occur in sequences of multiple marks. 
 
 The sets are:
-  - Below-base (class 220) MCMs
-  - Above-base (class 230) MCMs
+  - Below-base (class 220) <abbr>MCM</abbr>s
+  - Above-base (class 230) <abbr>MCM</abbr>s
   
 These classifications are used in the [mark-transient-reordering
 stage](#1-transient-reordering-of-modifier-combining-marks).
 
-Lists of the marks that belong to each MCM classes are included in the
+Lists of the marks that belong to each <abbr>MCM</abbr> classes are included in the
 script-specific shaping documents for Arabic and Syriac.
 			
 			
@@ -202,7 +202,7 @@ the general Arabic shaping model involves seven top-level stages:
 <!--- http://www.unicode.org/reports/tr53/tr53-1.pdf --->
 > Note: the transient reordering of modifier combining marks is
 > necessary only for scripts that can feature the "Shadda" mark or
-> marks that belong to _Modifier Combining Marks_ (MCM) classes.
+> marks that belong to _Modifier Combining Marks_ (<abbr>MCM</abbr>) classes.
 
 Sequences of adjacent marks must be reordered so that they appear in
 the appropriate visual order before the mark-to-base and mark-to-mark
@@ -216,7 +216,7 @@ cross-script mark-reordering performed during Unicode
 normalization. The standard Unicode mark-reordering algorithm is based
 on comparing the _Canonical_Combining_Class_ (Ccc) properties of mark
 codepoints, whereas this script-specific reordering utilizes the
-_Modifier_Combining_Mark_ (`MCM`) subclasses specified in the
+_Modifier_Combining_Mark_ (<abbr>MCM</abbr>) subclasses specified in the
 character tables.
 
 The algorithm for reordering a sequence of marks is:
@@ -235,11 +235,11 @@ The algorithm for reordering a sequence of marks is:
        characters. The subsequence must be moved as a group.
 
 > Note: Unicode describes this mark-reordering operation, the Arabic
-> Mark Transient Reordering Algorithm (AMTRA), in Technical Report 53,
+> Mark Transient Reordering Algorithm (<abbr>AMTRA</abbr>), in Technical Report 53,
 > which describes it in terms that are distinct from standard,
 > Ccc-based mark reordering.
 >
-> Specifically, AMTRA is designated as an operation performed during
+> Specifically, <abbr>AMTRA</abbr> is designated as an operation performed during
 > text rendering only, which therefore does not impact other
 > Unicode-compliance issues such as allowable input sequences or text
 > encoding.
@@ -379,7 +379,7 @@ Finally, the decomposed mark must be reordered as follows:
     the word.
 	
 
-### 5. Applying the language-form substitution features from GSUB ###
+### 5. Applying the language-form substitution features from <abbr>GSUB</abbr> ###
 
 The language-substitution phase applies mandatory substitution
 features using the rules in the font's <abbr>GSUB</abbr> table. In preparation for
@@ -416,7 +416,7 @@ for script-specific information.
 > <abbr>GSUB</abbr> substitutions in the following steps.
 
 
-### 6. Applying the typographic-form substitution features from GSUB ###
+### 6. Applying the typographic-form substitution features from <abbr>GSUB</abbr> ###
 
 The typographic-substitution phase applies optional substitution
 features using the rules in the font's <abbr>GSUB</abbr> table.
@@ -433,7 +433,7 @@ See the individual script pages for further detail on each feature and
 for script-specific information.
 
 
-### 7. Applying the positioning features from GPOS ###
+### 7. Applying the positioning features from <abbr>GPOS</abbr> ###
 
 The positioning stage adjusts the positions of mark and base
 glyphs.
diff --git a/opentype-shaping-arabic.md b/opentype-shaping-arabic.md
index de5daea..bfec75e 100644
--- a/opentype-shaping-arabic.md
+++ b/opentype-shaping-arabic.md
@@ -18,9 +18,9 @@ implementations share.
       - [2. Compound character composition and decomposition](#2-compound-character-composition-and-decomposition)
       - [3. Computing letter joining states](#3-computing-letter-joining-states)
       - [4. Applying the `stch` feature](#4-applying-the-stch-feature)
-      - [5. Applying the language-form substitution features from GSUB](#5-applying-the-language-form-substitution-features-from-gsub)
-      - [6. Applying the typographic-form substitution features from GSUB](#6-applying-the-typographic-form-substitution-features-from-gsub)
-      - [7. Applying the positioning features from GPOS](#7-applying-the-positioning-features-from-gpos)
+      - [5. Applying the language-form substitution features from <abbr>GSUB</abbr>](#5-applying-the-language-form-substitution-features-from-gsub)
+      - [6. Applying the typographic-form substitution features from <abbr>GSUB</abbr>](#6-applying-the-typographic-form-substitution-features-from-gsub)
+      - [7. Applying the positioning features from <abbr>GPOS</abbr>](#7-applying-the-positioning-features-from-gpos)
   
 
 
@@ -178,14 +178,14 @@ normalization.
 
 A subset of the Arabic marks require special handling when shaping
 Arabic text, during the mark-reordering stage. These include two sets
-of _Modifier Combining Marks_ (MCM) that may need to be repositioned
+of _Modifier Combining Marks_ (<abbr>MCM</abbr>) that may need to be repositioned
 closer to the base character, when they occur in sequences of multiple
 marks. 
 
 The sets are:
-  - Below-base (class 220) MCMs: "Hamza below" (`U+0655`), "Small low seen"
+  - Below-base (class 220) <abbr>MCM</abbr>s: "Hamza below" (`U+0655`), "Small low seen"
     (`U+06E3`), "Large round dot below" (`U+08CF`), "Small low waw" (`U+08D3`)
-  - Above-base (class 230) MCMs: "Hamza above" (`U+0654`), "Mark noon ghunna"
+  - Above-base (class 230) <abbr>MCM</abbr>s: "Hamza above" (`U+0654`), "Mark noon ghunna"
     (`U+0658`), "Small high seen" (`U+06DC`), "Small high yeh" (`U+06E7`), "Small high
     noon" (`U+06E8`), "Small high Farsi yeh" (`U+08CA`), "Small high
     yeh barree with two dots below" (`U+08CB`), "Small high zah"
@@ -286,17 +286,17 @@ This requirement covers:
 
 
 
-The combining grapheme joiner (CGJ) is primarily used to alter the
+The combining grapheme joiner (<abbr>CGJ</abbr>) is primarily used to alter the
 order in which adjacent marks are positioned during the
 mark-reordering stage, in order to adhere to the needs of a
 non-default language orthography.
 
 By default, OpenType shaping reorders sequences of adjacent marks by
 sorting the sequence on the marks' Canonical_Combining_Class (Ccc)
-values. The presence of a CGJ character within a sequence of marks has
+values. The presence of a <abbr>CGJ</abbr> character within a sequence of marks has
 the effect of splitting the sequence into two sequences of marks and,
 therefore, halting any mark-reordering that would have occurred
-between the marks on either side of the CGJ.
+between the marks on either side of the <abbr>CGJ</abbr>.
 
 The zero-width joiner (<abbr>ZWJ</abbr>) is primarily used to force the usage of the
 cursive connecting form of a letter even when the context of the
@@ -334,10 +334,10 @@ when matching sequences of codepoints against the backtrack and
 lookahead sequences of a font's <abbr>GSUB</abbr> or <abbr>GPOS</abbr> lookups.
 
 
-The right-to-left mark (RLM) and left-to-right mark (LRM) are used by
+The right-to-left mark (<abbr>RLM</abbr>) and left-to-right mark (<abbr>LRM</abbr>) are used by
 the Unicode bidirectionality algorithm (BiDi) to indicate the points
 in a text run at which the writing direction changes. Generally
-speaking RLM and LRM codepoints do not interact with shaping.
+speaking <abbr>RLM</abbr> and <abbr>LRM</abbr> codepoints do not interact with shaping.
 
 The no-break space is primarily used to display those codepoints that
 are defined as non-spacing (such as vowel or diacritical marks and "Hamza") in an
@@ -374,7 +374,7 @@ cross-script mark-reordering performed during Unicode
 normalization. The standard Unicode mark-reordering algorithm is based
 on comparing the _Canonical_Combining_Class_ (Ccc) properties of mark
 codepoints, whereas this script-specific reordering utilizes the
-_Modifier_Combining_Mark_ (`MCM`) subclasses specified in the
+_Modifier_Combining_Mark_ (<abbr>MCM</abbr>) subclasses specified in the
 character tables.
 
 The algorithm for reordering a sequence of marks is:
@@ -393,11 +393,11 @@ The algorithm for reordering a sequence of marks is:
        characters. The subsequence must be moved as a group.
 
 > Note: Unicode describes this mark-reordering operation, the Arabic
-> Mark Transient Reordering Algorithm (AMTRA), in Technical Report 53,
+> Mark Transient Reordering Algorithm (<abbr>AMTRA</abbr>), in Technical Report 53,
 > which describes it in terms that are distinct from standard,
 > Ccc-based mark reordering.
 >
-> Specifically, AMTRA is designated as an operation performed during
+> Specifically, <abbr>AMTRA</abbr> is designated as an operation performed during
 > text rendering only, which therefore does not impact other
 > Unicode-compliance issues such as allowable input sequences or text
 > encoding.
@@ -539,7 +539,7 @@ Finally, the decomposed mark must be reordered as follows:
     the word.
 	
 
-### 5. Applying the language-form substitution features from GSUB ###
+### 5. Applying the language-form substitution features from <abbr>GSUB</abbr> ###
 
 The language-substitution phase applies mandatory substitution
 features using the rules in the font's <abbr>GSUB</abbr> table. In preparation for
@@ -666,7 +666,7 @@ can be disabled by application-level user interfaces.
 
 
 
-### 6. Applying the typographic-form substitution features from GSUB ###
+### 6. Applying the typographic-form substitution features from <abbr>GSUB</abbr> ###
 
 The typographic-substitution phase applies optional substitution
 features using the rules in the font's <abbr>GSUB</abbr> table.
@@ -719,7 +719,7 @@ of bases and marks with precomposed base-and-mark glyphs.
 > Nevertheless, when the active font uses `mset` substitutions, the
 > shaping engine must deal with the situation gracefully.
 
-### 7. Applying the positioning features from GPOS ###
+### 7. Applying the positioning features from <abbr>GPOS</abbr> ###
 
 The positioning stage adjusts the positions of mark and base
 glyphs.
diff --git a/opentype-shaping-bengali.md b/opentype-shaping-bengali.md
index 6bc2cd2..aa39939 100644
--- a/opentype-shaping-bengali.md
+++ b/opentype-shaping-bengali.md
@@ -14,10 +14,10 @@ runs in the Bengali script.
   - [The `<bng2>` shaping model](#the-bng2-shaping-model)
       - [1: Identifying syllables and other sequences](#1-identifying-syllables-and-other-sequences)
       - [2: Initial reordering](#2-initial-reordering)
-      - [3: Applying the basic substitution features from GSUB](#3-applying-the-basic-substitution-features-from-gsub)
+      - [3: Applying the basic substitution features from <abbr>GSUB</abbr>](#3-applying-the-basic-substitution-features-from-gsub)
       - [4: Final reordering](#4-final-reordering)
-      - [5: Applying all remaining substitution features from GSUB](#5-applying-all-remaining-substitution-features-from-gsub)
-      - [6: Applying remaining positioning features from GPOS](#6-applying-remaining-positioning-features-from-gpos)
+      - [5: Applying all remaining substitution features from <abbr>GSUB</abbr>](#5-applying-all-remaining-substitution-features-from-gsub)
+      - [6: Applying remaining positioning features from <abbr>GPOS</abbr>](#6-applying-remaining-positioning-features-from-gpos)
   - [The `<beng>` shaping model](#the-beng-shaping-model)
       - [Distinctions from `<bng2>`](#distinctions-from-bng2)
       - [Advice for handling fonts with `<beng>` features only](#advice-for-handling-fonts-with-beng-features-only)
@@ -328,7 +328,7 @@ For example:
     should still be applied if the dependent-vowel codepoint is preceded
     by "Ka,Halant,ZWJ,Tta" in the text run.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
@@ -1193,7 +1193,7 @@ in <abbr>GSUB</abbr> and <abbr>GPOS</abbr> application stages that follow.
 
 
 
-### 3: Applying the basic substitution features from GSUB ###
+### 3: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -1641,7 +1641,7 @@ of <abbr>GSUB</abbr>.
 > note](/notes/uniscribe-bug-compatibility.md). 
 
 
-### 5: Applying all remaining substitution features from GSUB ###
+### 5: Applying all remaining substitution features from <abbr>GSUB</abbr> ###
 
 In this stage, the remaining substitution features from the <abbr>GSUB</abbr> table
 are applied. In preparation for this stage, glyph sequences should be
@@ -1716,7 +1716,7 @@ typographically problematic.
 > point. However, `calt` is not mandatory for correct Bengali shaping
 > and may be disabled in the application by user preference.
 
-### 6: Applying remaining positioning features from GPOS ###
+### 6: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied.
diff --git a/opentype-shaping-default.md b/opentype-shaping-default.md
index 7a23a00..b488b55 100644
--- a/opentype-shaping-default.md
+++ b/opentype-shaping-default.md
@@ -11,9 +11,9 @@ model for unrecognized scripts.
   - [Terminology](#terminology)
   - [Normalization](#normalization)
   - [The default shaping model](#the-default-shaping-model)
-      - [1: Applying the basic substitution features from GSUB](#1-applying-the-basic-substitution-features-from-gsub)
-	  - [2: Applying typographic substitution features from GSUB](#2-applying-typographic-substitution-features-from-gsub)
-	  - [3: Applying the positioning features from GPOS](#3-applying-the-positioning-features-from-gpos)
+      - [1: Applying the basic substitution features from <abbr>GSUB</abbr>](#1-applying-the-basic-substitution-features-from-gsub)
+	  - [2: Applying typographic substitution features from <abbr>GSUB</abbr>](#2-applying-typographic-substitution-features-from-gsub)
+	  - [3: Applying the positioning features from <abbr>GPOS</abbr>](#3-applying-the-positioning-features-from-gpos)
   
   
   
@@ -123,7 +123,7 @@ for processing, or may choose to reply on higher-level applications to
 make segmentation decisions.
 
 
-### 1: Applying the basic substitution features from GSUB ###
+### 1: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -170,7 +170,7 @@ ligatures. Substitutions made by `rlig` cannot be disabled by
 application-level user interfaces.
 
 
-### 2: Applying typographic substitution features from GSUB ###
+### 2: Applying typographic substitution features from <abbr>GSUB</abbr> ###
 
 The typographic-substitution phase applies all remaining substitution
 features using the rules in the font's <abbr>GSUB</abbr> table. In preparation for
@@ -213,7 +213,7 @@ by default. Substitutions made by `liga` may be disabled by
 application-level user interfaces.
 
 
-### 3: Applying the positioning features from GPOS ###
+### 3: Applying the positioning features from <abbr>GPOS</abbr> ###
 
 The positioning stage adjusts the positions of mark and base
 glyphs. In preparation for this stage, glyph sequences should be
diff --git a/opentype-shaping-devanagari.md b/opentype-shaping-devanagari.md
index 1202530..afce0e9 100644
--- a/opentype-shaping-devanagari.md
+++ b/opentype-shaping-devanagari.md
@@ -14,10 +14,10 @@ runs in the Devanagari script.
   - [The `<dev2>` shaping model](#the-dev2-shaping-model)
       - [1: Identifying syllables and other sequences](#1-identifying-syllables-and-other-sequences)
       - [2: Initial reordering](#2-initial-reordering)
-      - [3: Applying the basic substitution features from GSUB](#3-applying-the-basic-substitution-features-from-gsub)
+      - [3: Applying the basic substitution features from <abbr>GSUB</abbr>](#3-applying-the-basic-substitution-features-from-gsub)
       - [4: Final reordering](#4-final-reordering)
-      - [5: Applying all remaining substitution features from GSUB](#5-applying-all-remaining-substitution-features-from-gsub)
-      - [6: Applying remaining positioning features from GPOS](#6-applying-remaining-positioning-features-from-gpos)
+      - [5: Applying all remaining substitution features from <abbr>GSUB</abbr>](#5-applying-all-remaining-substitution-features-from-gsub)
+      - [6: Applying remaining positioning features from <abbr>GPOS</abbr>](#6-applying-remaining-positioning-features-from-gpos)
   - [The `<deva>` shaping model](#the-deva-shaping-model)
       - [Distinctions from `<dev2>`](#distinctions-from-dev2)
       - [Advice for handling fonts with `<deva>` features only](#advice-for-handling-fonts-with-deva-features-only)
@@ -319,7 +319,7 @@ For example:
     should still be applied if the dependent-vowel codepoint is preceded
     by "Ka,Halant,ZWJ,Tta" in the text run.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
@@ -1114,7 +1114,7 @@ in <abbr>GSUB</abbr> and <abbr>GPOS</abbr> application stages that follow.
 
 
 
-### 3: Applying the basic substitution features from GSUB ###
+### 3: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -1535,7 +1535,7 @@ involve no work when processing `<dev2>` text. It is included here in
 order to maintain compatibility with the other Indic scripts.
 
 
-### 5: Applying all remaining substitution features from GSUB ###
+### 5: Applying all remaining substitution features from <abbr>GSUB</abbr> ###
 
 In this stage, the remaining substitution features from the <abbr>GSUB</abbr> table
 are applied. In preparation for this stage, glyph sequences should be
@@ -1594,7 +1594,7 @@ typographically problematic.
 > point. However, `calt` is not mandatory for correct Devanagari shaping
 > and may be disabled in the application by user preference.
 
-### 6: Applying remaining positioning features from GPOS ###
+### 6: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied.
diff --git a/opentype-shaping-emoji.md b/opentype-shaping-emoji.md
index 58b63b9..fbb88e6 100644
--- a/opentype-shaping-emoji.md
+++ b/opentype-shaping-emoji.md
@@ -17,13 +17,13 @@ emoji sequences.
     - [Tag flag sequences](#tag-flag-sequences)
     - [Keycap sequences](#keycap-sequences)
     - [Zero-Width Joiner sequences](#zwj-sequences)
-      - [ZWJ hair sequences](#zwj-hair-sequences)
-      - [ZWJ gendered person sequences](#zwj-gendered-person-sequences)
-      - [ZWJ multi-person group sequences](#zwj-multi-person-group-sequences)
-      - [ZWJ role sequences](#zwj-role-sequences)
-      - [ZWJ color sequences](#zwj-color-sequences)
-      - [ZWJ directionality sequences](#zwj-directionality-sequences)
-      - [ZWJ additional sequences](#zwj-additional-sequences)
+      - [<abbr>ZWJ</abbr> hair sequences](#zwj-hair-sequences)
+      - [<abbr>ZWJ</abbr> gendered person sequences](#zwj-gendered-person-sequences)
+      - [<abbr>ZWJ</abbr> multi-person group sequences](#zwj-multi-person-group-sequences)
+      - [<abbr>ZWJ</abbr> role sequences](#zwj-role-sequences)
+      - [<abbr>ZWJ</abbr> color sequences](#zwj-color-sequences)
+      - [<abbr>ZWJ</abbr> directionality sequences](#zwj-directionality-sequences)
+      - [<abbr>ZWJ</abbr> additional sequences](#zwj-additional-sequences)
     - [Other sequences and ligatures](#other-sequences-and-ligatures)
   - [Feature interaction in sequences](#feature-interaction-in-sequences)
   - [Emoji sets](#emoji-sets)
@@ -96,7 +96,7 @@ emoji.
 ## Terminology ##
 
 A codepoint is considered an **emoji** only if it has the `Emoji`
-property in the Unicode Character Database (UCD). Although many
+property in the Unicode Character Database (<abbr>UCD</abbr>). Although many
 codepoints that have this property are pictographic in nature, some
 codepoints that are pictographic do not have the `Emoji` property
 (such as most chess, playing-card, and game-piece symbols), and some
@@ -455,7 +455,7 @@ environment.
 Regional Indicator flag sequences only occur in standalone form.
 
 > Note: The Regional Indicator flag sequences are defined to always be
-> interpreted left-to-right (LTR) for the purpose of
+> interpreted left-to-right (<abbr>LTR</abbr>) for the purpose of
 > bidirectionality. This behavior differs from that of other emoji
 > sequences, which are neutral in regard to bidirectionality.
 >
@@ -511,7 +511,7 @@ Keycap sequences only occur in standalone form.
 
 
 
-### ZWJ sequences ###
+### <abbr>ZWJ</abbr> sequences ###
 
 A Zero-Width Joiner (<abbr>ZWJ</abbr>) sequence can be used to request specific
 variants of an emoji glyph or to request the combined form of a
@@ -527,7 +527,7 @@ back to a set of individual-person emoji glyphs.
 
 
 
-#### ZWJ hair sequences ####
+#### <abbr>ZWJ</abbr> hair sequences ####
 
 A <abbr>ZWJ</abbr> hair sequence is used to request a specific hairstyle version of
 an emoji codepoint that depicts a single human being.
@@ -558,7 +558,7 @@ therefore "blond" was not included in the set of supported hairstyle
 versions.
 
 
-#### ZWJ gendered person sequences ####
+#### <abbr>ZWJ</abbr> gendered person sequences ####
 
 A <abbr>ZWJ</abbr> gendered person sequence is used to request a specific-gendered
 version of an emoji codepoint that depicts a single human being.
@@ -625,7 +625,7 @@ group mechanism, and are documented in the corresponding section.
 
 
 
-#### ZWJ multi-person group sequences ####
+#### <abbr>ZWJ</abbr> multi-person group sequences ####
 
 A <abbr>ZWJ</abbr> multi-person group sequence is used to request a multi-person
 emoji glyph. The fallback for a <abbr>ZWJ</abbr> multi-person group sequence is a
@@ -750,7 +750,7 @@ separated by a <abbr>ZWJ</abbr>, and must match:
 > group sequence for "Couple holding hands".
 
 
-#### ZWJ role sequences ####
+#### <abbr>ZWJ</abbr> role sequences ####
 
 A <abbr>ZWJ</abbr> role (or profession) sequence is used to request an emoji
 depicting a human being performing a task or job. Role sequences are
@@ -804,7 +804,7 @@ _genderperson_ _modifier_? _zwj_ _emoji_ _presentation_?
 
 
 
-#### ZWJ color sequences ####
+#### <abbr>ZWJ</abbr> color sequences ####
 
 A <abbr>ZWJ</abbr> color sequence is used to request a version of an emoji
 codepoint depicting the base object in a specific color.
@@ -836,7 +836,7 @@ the default emoji for the color codepoint (that is, the color square).
 
 
 
-#### ZWJ directionality sequences ####
+#### <abbr>ZWJ</abbr> directionality sequences ####
 
 A <abbr>ZWJ</abbr> directionality sequence is used to request a version of an emoji
 codepoint facing a specific cardinal direction.
@@ -850,7 +850,7 @@ A <abbr>ZWJ</abbr> directionality sequence must match:
 _emoji_ _zwj_ _direction_ _presentation_
 ```
 
-#### ZWJ additional sequences ####
+#### <abbr>ZWJ</abbr> additional sequences ####
 
 In addition to the above <abbr>ZWJ</abbr> sequence categories, there are 13
 standalone, but uncategorized, <abbr>ZWJ</abbr> sequences defined in Unicode.
@@ -968,7 +968,7 @@ General Interchange" (<abbr>RGI</abbr>) to denote that they are in common usage.
 
 Finally, the "<abbr>RGI</abbr> emoji set" includes all of the codepoints and
 sequences included in the preceding sets. Presence in the <abbr>RGI</abbr> emoji
-set can be tracked with the `RGI_Emoji` property in the UCD. Fonts are
+set can be tracked with the `RGI_Emoji` property in the <abbr>UCD</abbr>. Fonts are
 not required to implement the entire <abbr>RGI</abbr> emoji set, nor any of the
 other sets.
 
@@ -1022,7 +1022,7 @@ require font substitution.
 > and handle them gracefully.
 
 
-### 1. Applying the basic substitution features from GSUB ###
+### 1. Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -1078,7 +1078,7 @@ shaping process. Emoji shaping should be unaffected by this decision.
 
 
 
-### 2. Applying typographic substitution features from GSUB ###
+### 2. Applying typographic substitution features from <abbr>GSUB</abbr> ###
 
 The typographic-substitution phase applies all remaining substitution
 features using the rules in the font's <abbr>GSUB</abbr> table. In preparation for
@@ -1114,7 +1114,7 @@ in `rlig`, with custom sequences implemented in `liga`.
 
 
 
-### 3. Applying the positioning features from GPOS ###
+### 3. Applying the positioning features from <abbr>GPOS</abbr> ###
 
 The positioning stage adjusts the positions of mark and base
 glyphs. In preparation for this stage, glyph sequences should be
diff --git a/opentype-shaping-gujarati.md b/opentype-shaping-gujarati.md
index 3900f38..3f549eb 100644
--- a/opentype-shaping-gujarati.md
+++ b/opentype-shaping-gujarati.md
@@ -14,10 +14,10 @@ runs in the Gujarati script.
   - [The `<gjr2>` shaping model](#the-gjr2-shaping-model)
       - [1: Identifying syllables and other sequences](#1-identifying-syllables-and-other-sequences)
       - [2: Initial reordering](#2-initial-reordering)
-      - [3: Applying the basic substitution features from GSUB](#3-applying-the-basic-substitution-features-from-gsub)
+      - [3: Applying the basic substitution features from <abbr>GSUB</abbr>](#3-applying-the-basic-substitution-features-from-gsub)
       - [4: Final reordering](#4-final-reordering)
-      - [5: Applying all remaining substitution features from GSUB](#5-applying-all-remaining-substitution-features-from-gsub)
-      - [6: Applying remaining positioning features from GPOS](#6-applying-remaining-positioning-features-from-gpos)
+      - [5: Applying all remaining substitution features from <abbr>GSUB</abbr>](#5-applying-all-remaining-substitution-features-from-gsub)
+      - [6: Applying remaining positioning features from <abbr>GPOS</abbr>](#6-applying-remaining-positioning-features-from-gpos)
   - [The `<gujr>` shaping model](#the-gujr-shaping-model)
       - [Distinctions from `<gjr2>`](#distinctions-from-gjr2)
       - [Advice for handling fonts with `<gujr>` features only](#advice-for-handling-fonts-with-gujr-features-only)
@@ -314,7 +314,7 @@ For example:
     should still be applied if the dependent-vowel codepoint is preceded
     by "Ka,Halant,ZWJ,Tta" in the text run.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
@@ -1107,7 +1107,7 @@ in <abbr>GSUB</abbr> and <abbr>GPOS</abbr> application stages that follow.
 
 
 
-### 3: Applying the basic substitution features from GSUB ###
+### 3: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -1528,7 +1528,7 @@ involve no work when processing `<gjr2>` text. It is included here in
 order to maintain compatibility with the other Indic scripts.
 
    
-### 5: Applying all remaining substitution features from GSUB ###
+### 5: Applying all remaining substitution features from <abbr>GSUB</abbr> ###
 
 In this stage, the remaining substitution features from the <abbr>GSUB</abbr> table
 are applied. In preparation for this stage, glyph sequences should be
@@ -1588,7 +1588,7 @@ typographically problematic.
 > point. However, `calt` is not mandatory for correct Gujarati shaping
 > and may be disabled in the application by user preference.
 
-### 6: Applying remaining positioning features from GPOS ###
+### 6: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied.
diff --git a/opentype-shaping-gurmukhi.md b/opentype-shaping-gurmukhi.md
index 7d2f99d..43589f4 100644
--- a/opentype-shaping-gurmukhi.md
+++ b/opentype-shaping-gurmukhi.md
@@ -14,10 +14,10 @@ runs in the Gurmukhi script.
   - [The `<gur2>` shaping model](#the-gur2-shaping-model)
       - [1: Identifying syllables and other sequences](#1-identifying-syllables-and-other-sequences)
       - [2: Initial reordering](#2-initial-reordering)
-      - [3: Applying the basic substitution features from GSUB](#3-applying-the-basic-substitution-features-from-gsub)
+      - [3: Applying the basic substitution features from <abbr>GSUB</abbr>](#3-applying-the-basic-substitution-features-from-gsub)
       - [4: Final reordering](#4-final-reordering)
-      - [5: Applying all remaining substitution features from GSUB](#5-applying-all-remaining-substitution-features-from-gsub)
-      - [6: Applying remaining positioning features from GPOS](#6-applying-remaining-positioning-features-from-gpos)
+      - [5: Applying all remaining substitution features from <abbr>GSUB</abbr>](#5-applying-all-remaining-substitution-features-from-gsub)
+      - [6: Applying remaining positioning features from <abbr>GPOS</abbr>](#6-applying-remaining-positioning-features-from-gpos)
   - [The `<guru>` shaping model](#the-guru-shaping-model)
       - [Distinctions from `<gur2>`](#distinctions-from-gur2)
       - [Advice for handling fonts with `<guru>` features only](#advice-for-handling-fonts-with-guru-features-only)
@@ -357,7 +357,7 @@ For example:
     should still be applied if the dependent-vowel codepoint is preceded
     by "Ka,Halant,ZWJ,Tta" in the text run.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
@@ -1176,7 +1176,7 @@ in <abbr>GSUB</abbr> and <abbr>GPOS</abbr> application stages that follow.
 
 
 
-### 3: Applying the basic substitution features from GSUB ###
+### 3: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -1601,7 +1601,7 @@ involve no work when processing `<gur2>` text. It is included here in
 order to maintain compatibility with the other Indic scripts.
 
 
-### 5: Applying all remaining substitution features from GSUB ###
+### 5: Applying all remaining substitution features from <abbr>GSUB</abbr> ###
 
 In this stage, the remaining substitution features from the <abbr>GSUB</abbr> table
 are applied. In preparation for this stage, glyph sequences should be
@@ -1660,7 +1660,7 @@ typographically problematic.
 > point. However, `calt` is not mandatory for correct Gurmukhi shaping
 > and may be disabled in the application by user preference.
 
-### 6: Applying remaining positioning features from GPOS ###
+### 6: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied.
diff --git a/opentype-shaping-hangul.md b/opentype-shaping-hangul.md
index 96068c7..204c0a1 100644
--- a/opentype-shaping-hangul.md
+++ b/opentype-shaping-hangul.md
@@ -18,7 +18,7 @@ implementations share.
       - [2. Determining if the syllable can be composed into a Hangul Syllables codepoint](#2-determining-if-the-syllable-can-be-composed-into-a-hangul-syllables-codepoint)
       - [3. Composing the syllable (if composition is possible)](#3-composing-the-syllable-if-composition-is-possible)
       - [4. Fully decomposing the syllable (if composition is not possible)](#4-fully-decomposing-the-syllable-if-composition-is-not-possible)
-      - [5. Shaping the fully decomposed syllable with GSUB features](#5-shaping-the-fully-decomposed-syllable-with-gsub-features)
+      - [5. Shaping the fully decomposed syllable with <abbr>GSUB</abbr> features](#5-shaping-the-fully-decomposed-syllable-with-gsub-features)
       - [6. Reordering tone marks](#6-reordering-tone-marks)
  
 
@@ -475,7 +475,7 @@ five with the `L`, `V`, and (if used) `T` jamo.
 
 
 
-### 5. Shaping the fully decomposed syllable with GSUB features ###
+### 5. Shaping the fully decomposed syllable with <abbr>GSUB</abbr> features ###
 
 With the syllable fully decomposed into a sequence of jamo, the next
 stage applies mandatory substitution features using rules in the
diff --git a/opentype-shaping-hebrew.md b/opentype-shaping-hebrew.md
index d4a42f6..104b9e0 100644
--- a/opentype-shaping-hebrew.md
+++ b/opentype-shaping-hebrew.md
@@ -15,9 +15,9 @@ implementations share.
   - [The `<hebr>` shaping model](#the-hebr-shaping-model)
       - [1. Compound character composition and decomposition](#1-compound-character-composition-and-decomposition)
       - [2. Composing any Alphabetic Presentation forms](#2-composing-any-alphabetic-presentation-forms)
-      - [3. Applying the language-form substitution features from GSUB](#3-applying-the-language-form-substitution-features-from-gsub)
-      - [4. Applying the typographic-form substitution features from GSUB](#4-applying-the-typographic-form-substitution-features-from-gsub)
-      - [5. Applying the positioning features from GPOS](#5-applying-the-positioning-features-from-gpos)
+      - [3. Applying the language-form substitution features from <abbr>GSUB</abbr>](#3-applying-the-language-form-substitution-features-from-gsub)
+      - [4. Applying the typographic-form substitution features from <abbr>GSUB</abbr>](#4-applying-the-typographic-form-substitution-features-from-gsub)
+      - [5. Applying the positioning features from <abbr>GPOS</abbr>](#5-applying-the-positioning-features-from-gpos)
   
 
 
@@ -192,7 +192,7 @@ also use other characters, such as hyphens or dashes, in a similar
 placeholder fashion; shaping engines should cope with this situation
 gracefully.
 
-The combining grapheme joiner (CGJ), zero-width joiner (<abbr>ZWJ</abbr>), and
+The combining grapheme joiner (<abbr>CGJ</abbr>), zero-width joiner (<abbr>ZWJ</abbr>), and
 zero-width non-joiner (<abbr>ZWNJ</abbr>) may be used to alter the
 order in which adjacent marks are positioned during the
 mark-reordering stage, in order to adhere to the needs of a
@@ -203,12 +203,12 @@ non-default language orthography.
 
 <!--- Zero-Width Non Joiner explanation --->
 
-The right-to-left mark (RLM) and left-to-right mark (LRM) are used by
+The right-to-left mark (<abbr>RLM</abbr>) and left-to-right mark (<abbr>LRM</abbr>) are used by
 the Unicode bidirectionality algorithm (BiDi) to indicate the points
 in a text run at which the writing direction changes.
 
 
-<!--- How shaping is affected by the LTR and RTL markers explanation --->
+<!--- How shaping is affected by the <abbr>LTR</abbr> and <abbr>RTL</abbr> markers explanation --->
 
 
 The no-break space may be used to display those codepoints that
@@ -305,7 +305,7 @@ glyphs in the block have a single composing sequence.
 ![Alphabetic Presentation forms composition](/images/hebrew/hebrew-apf.png)
 
 
-### 3. Applying the language-form substitution features from GSUB ###
+### 3. Applying the language-form substitution features from <abbr>GSUB</abbr> ###
 
 The language-substitution phase applies mandatory substitution
 features using the rules in the font's <abbr>GSUB</abbr> table. In preparation for
@@ -332,7 +332,7 @@ variants, based on examining the language setting of the text run.
 
 
 
-### 4. Applying the typographic-form substitution features from GSUB ###
+### 4. Applying the typographic-form substitution features from <abbr>GSUB</abbr> ###
 
 The typographic-substitution phase applies optional substitution
 features using the rules in the font's <abbr>GSUB</abbr> table.
@@ -364,7 +364,7 @@ application-level user interfaces.
 
 
 
-### 5. Applying the positioning features from GPOS ###
+### 5. Applying the positioning features from <abbr>GPOS</abbr> ###
 
 The positioning stage adjusts the positions of mark and base
 glyphs.
diff --git a/opentype-shaping-indic-general.md b/opentype-shaping-indic-general.md
index f34f0a0..280e8a2 100644
--- a/opentype-shaping-indic-general.md
+++ b/opentype-shaping-indic-general.md
@@ -18,10 +18,10 @@ implementations share.
       - [Script shaping characteristics](#script-shaping-characteristics)
       - [1: Identifying syllables and other sequences](#1-identifying-syllables-and-other-sequences)
       - [2: Initial reordering](#2-initial-reordering)
-      - [3: Applying the basic substitution features from GSUB](#3-applying-the-basic-substitution-features-from-gsub)
+      - [3: Applying the basic substitution features from <abbr>GSUB</abbr>](#3-applying-the-basic-substitution-features-from-gsub)
       - [4: Final reordering](#4-final-reordering)
-      - [5: Applying all remaining substitution features from GSUB](#5-applying-all-remaining-substitution-features-from-gsub)
-      - [6: Applying remaining positioning features from GPOS](#6-applying-remaining-positioning-features-from-gpos)
+      - [5: Applying all remaining substitution features from <abbr>GSUB</abbr>](#5-applying-all-remaining-substitution-features-from-gsub)
+      - [6: Applying remaining positioning features from <abbr>GPOS</abbr>](#6-applying-remaining-positioning-features-from-gpos)
   - [The old Indic shaping model](#the-old-indic-shaping-model)
 
 
@@ -371,7 +371,7 @@ For example:
     should still be applied if the dependent-vowel codepoint is preceded
     by "Ka,Halant,ZWJ,Tta" in the text run.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
@@ -408,7 +408,7 @@ converted into a combining mark-like form.
 
 The resulting mark must be correctly positioned by attaching it to the
 correct base character using the active font's `mark` lookup from
-`<abbr>GPOS</abbr>`. Therefore, the mark form of the "Ra" must be moved so that it
+<abbr>GPOS</abbr>. Therefore, the mark form of the "Ra" must be moved so that it
 is adjacent to the correct base character. Which character in a
 syllable is the correct base character differs from script to script,
 and may involve several context-sensitive tests.
@@ -1430,7 +1430,7 @@ in <abbr>GSUB</abbr> and <abbr>GPOS</abbr> application stages that follow.
 
 
 
-### 3: Applying the basic substitution features from GSUB ###
+### 3: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -1633,7 +1633,7 @@ of <abbr>GSUB</abbr>.
 > processing non-`<bng2>` text. 
 
 
-### 5: Applying all remaining substitution features from GSUB ###
+### 5: Applying all remaining substitution features from <abbr>GSUB</abbr> ###
 
 In this stage, the remaining substitution features from the <abbr>GSUB</abbr> table
 are applied. In preparation for this stage, glyph sequences should be
@@ -1651,7 +1651,7 @@ in the font.
 	psts
 	haln
 
-### 6: Applying remaining positioning features from GPOS ###
+### 6: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied.
diff --git a/opentype-shaping-kannada.md b/opentype-shaping-kannada.md
index 3c4484e..f5cbccf 100644
--- a/opentype-shaping-kannada.md
+++ b/opentype-shaping-kannada.md
@@ -14,10 +14,10 @@ runs in the Kannada script.
   - [The `<knd2>` shaping model](#the-knd2-shaping-model)
       - [1: Identifying syllables and other sequences](#1-identifying-syllables-and-other-sequences)
       - [2: Initial reordering](#2-initial-reordering)
-      - [3: Applying the basic substitution features from GSUB](#3-applying-the-basic-substitution-features-from-gsub)
+      - [3: Applying the basic substitution features from <abbr>GSUB</abbr>](#3-applying-the-basic-substitution-features-from-gsub)
       - [4: Final reordering](#4-final-reordering)
-      - [5: Applying all remaining substitution features from GSUB](#5-applying-all-remaining-substitution-features-from-gsub)
-      - [6: Applying remaining positioning features from GPOS](#6-applying-remaining-positioning-features-from-gpos)
+      - [5: Applying all remaining substitution features from <abbr>GSUB</abbr>](#5-applying-all-remaining-substitution-features-from-gsub)
+      - [6: Applying remaining positioning features from <abbr>GPOS</abbr>](#6-applying-remaining-positioning-features-from-gpos)
   - [The `<knda>` shaping model](#the-knda-shaping-model)
       - [Distinctions from `<knd2>`](#distinctions-from-knd2)
       - [Advice for handling fonts with `<knda>` features only](#advice-for-handling-fonts-with-knda-features-only)
@@ -333,7 +333,7 @@ For example:
     should still be applied if the dependent-vowel codepoint is preceded
     by "Ka,Halant,ZWJ,Tta" in the text run.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
@@ -1147,7 +1147,7 @@ in <abbr>GSUB</abbr> and <abbr>GPOS</abbr> application stages that follow.
 
 
 
-### 3: Applying the basic substitution features from GSUB ###
+### 3: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -1565,7 +1565,7 @@ typographically problematic.
 > point. However, `calt` is not mandatory for correct Kannada shaping
 > and may be disabled in the application by user preference.
 
-### 6: Applying remaining positioning features from GPOS ###
+### 6: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied.
diff --git a/opentype-shaping-khmer.md b/opentype-shaping-khmer.md
index 9157102..504ce41 100644
--- a/opentype-shaping-khmer.md
+++ b/opentype-shaping-khmer.md
@@ -14,9 +14,9 @@ runs in the Khmer script.
   - [The `<khmr>` shaping model](#the-khmr-shaping-model)
       - [1: Identifying syllables and other sequences](#1-identifying-syllables-and-other-sequences)
       - [2: Initial reordering](#2-initial-reordering)
-      - [3: Applying the basic substitution features from GSUB](#3-applying-the-basic-substitution-features-from-gsub)
-      - [4: Applying all remaining substitution features from GSUB](#4-applying-all-remaining-substitution-features-from-gsub)
-      - [5: Applying remaining positioning features from GPOS](#5-applying-remaining-positioning-features-from-gpos)
+      - [3: Applying the basic substitution features from <abbr>GSUB</abbr>](#3-applying-the-basic-substitution-features-from-gsub)
+      - [4: Applying all remaining substitution features from <abbr>GSUB</abbr>](#4-applying-all-remaining-substitution-features-from-gsub)
+      - [5: Applying remaining positioning features from <abbr>GPOS</abbr>](#5-applying-remaining-positioning-features-from-gpos)
 
 
 ## General information ##
@@ -308,12 +308,13 @@ where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 --->
 
-The no-break space is primarily used to display those codepoints that
-are defined as non-spacing (marks, dependent vowels (matras),
-below-base consonant forms, and post-base consonant forms) in an
-isolated context, as an alternative to displaying them superimposed on
-the dotted-circle placeholder. These sequences will match
-"NBSP,ZWJ,Sign_Coeng,_Consonant_", "NBSP,_mark_", or "NBSP,_matra_".
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display
+those codepoints that are defined as non-spacing (marks, dependent
+vowels (matras), below-base consonant forms, and post-base consonant
+forms) in an isolated context, as an alternative to displaying them
+superimposed on the dotted-circle placeholder. These sequences will
+match "NBSP,ZWJ,Sign_Coeng,_Consonant_", "NBSP,_mark_", or
+"NBSP,_matra_".
 
 The zero-width space may be used between words — even though no visual
 word spacing results — in order to indicate word breaks within a text
@@ -808,7 +809,7 @@ any consonants after the base consonant, the base consonant should
 With these steps completed, the syllable can be sorted into the final sort order.
 
 
-### 3: Applying the basic substitution features from GSUB ###
+### 3: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -918,7 +919,7 @@ pre-base-reordering "Ro".
 <!--- ### 4: Final reordering ### --->
 <!--- Is there any? --->
 
-### 4. Applying all remaining substitution features from GSUB ###
+### 4. Applying all remaining substitution features from <abbr>GSUB</abbr> ###
 
 In this stage, the remaining substitution features from the <abbr>GSUB</abbr> table
 are applied. The order in which these features are applied is not
@@ -984,7 +985,7 @@ application-level user interfaces.
 
 
 
-### 5: Applying remaining positioning features from GPOS ###
+### 5: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied. As with the preceding stage, the order in which these
diff --git a/opentype-shaping-malayalam.md b/opentype-shaping-malayalam.md
index b66e612..e07ec45 100644
--- a/opentype-shaping-malayalam.md
+++ b/opentype-shaping-malayalam.md
@@ -14,10 +14,10 @@ runs in the Malayalam script.
   - [The `<mlm2>` shaping model](#the-mlm2-shaping-model)
       - [1: Identifying syllables and other sequences](#1-identifying-syllables-and-other-sequences)
       - [2: Initial reordering](#2-initial-reordering)
-      - [3: Applying the basic substitution features from GSUB](#3-applying-the-basic-substitution-features-from-gsub)
+      - [3: Applying the basic substitution features from <abbr>GSUB</abbr>](#3-applying-the-basic-substitution-features-from-gsub)
       - [4: Final reordering](#4-final-reordering)
-      - [5: Applying all remaining substitution features from GSUB](#5-applying-all-remaining-substitution-features-from-gsub)
-      - [6: Applying remaining positioning features from GPOS](#6-applying-remaining-positioning-features-from-gpos)
+      - [5: Applying all remaining substitution features from <abbr>GSUB</abbr>](#5-applying-all-remaining-substitution-features-from-gsub)
+      - [6: Applying remaining positioning features from <abbr>GPOS</abbr>](#6-applying-remaining-positioning-features-from-gpos)
   - [The `<mlym>` shaping model](#the-mlym-shaping-model)
       - [Distinctions from `<mlm2>`](#distinctions-from-mlm2)
       - [Advice for handling fonts with `<mlym>` features only](#advice-for-handling-fonts-with-mlym-features-only)
@@ -348,7 +348,7 @@ For example:
     should still be applied if the dependent-vowel codepoint is preceded
     by "Ka,Halant,ZWJ,Tta" in the text run.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
@@ -1176,7 +1176,7 @@ in <abbr>GSUB</abbr> and <abbr>GPOS</abbr> application stages that follow.
 
 
 
-### 3: Applying the basic substitution features from GSUB ###
+### 3: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -1599,7 +1599,7 @@ involve no work when processing `<mlm2>` text. It is included here in
 order to maintain compatibility with the other Indic scripts.
 
 
-### 5: Applying all remaining substitution features from GSUB ###
+### 5: Applying all remaining substitution features from <abbr>GSUB</abbr> ###
 
 In this stage, the remaining substitution features from the <abbr>GSUB</abbr> table
 are applied. In preparation for this stage, glyph sequences should be
@@ -1656,7 +1656,7 @@ typographically problematic.
 > point. However, `calt` is not mandatory for correct Malayalam shaping
 > and may be disabled in the application by user preference.
 
-### 6: Applying remaining positioning features from GPOS ###
+### 6: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied.
diff --git a/opentype-shaping-mongolian.md b/opentype-shaping-mongolian.md
index abb8e84..a925d14 100644
--- a/opentype-shaping-mongolian.md
+++ b/opentype-shaping-mongolian.md
@@ -18,9 +18,9 @@ implementations share.
       - [2. Compound character composition and decomposition](#2-compound-character-composition-and-decomposition)
       - [3. Computing letter joining states](#3-computing-letter-joining-states)
       - [4. Applying the `stch` feature](#4-applying-the-stch-feature)
-      - [5. Applying the language-form substitution features from GSUB](#5-applying-the-language-form-substitution-features-from-gsub)
-      - [6. Applying the typographic-form substitution features from GSUB](#6-applying-the-typographic-form-substitution-features-from-gsub)
-      - [7. Applying the positioning features from GPOS](#7-applying-the-positioning-features-from-gpos)
+      - [5. Applying the language-form substitution features from <abbr>GSUB</abbr>](#5-applying-the-language-form-substitution-features-from-gsub)
+      - [6. Applying the typographic-form substitution features from <abbr>GSUB</abbr>](#6-applying-the-typographic-form-substitution-features-from-gsub)
+      - [7. Applying the positioning features from <abbr>GPOS</abbr>](#7-applying-the-positioning-features-from-gpos)
   
 
 
@@ -337,7 +337,7 @@ Processing a run of `<mong>` text involves seven top-level stages:
 
 <!--- http://www.unicode.org/reports/tr53/tr53-1.pdf --->
 > Note: because Mongolian does not feature the "Shadda" mark or any
-> marks that belong to _Modifier Combining Marks_ (MCM) classes, this
+> marks that belong to _Modifier Combining Marks_ (<abbr>MCM</abbr>) classes, this
 > stage should not involve any additional work when processing
 > `<mong>` text runs. It is included here to maintain consistency with
 > other scripts that utilize the general Arabic-based shaping model.
@@ -354,7 +354,7 @@ cross-script mark-reordering performed during Unicode
 normalization. The standard Unicode mark-reordering algorithm is based
 on comparing the _Canonical_Combining_Class_ (Ccc) properties of mark
 codepoints, whereas this script-specific reordering utilizes the
-_Modifier_Combining_Mark_ (`MCM`) subclasses specified in the
+_Modifier_Combining_Mark_ (<abbr>MCM</abbr>) subclasses specified in the
 character tables.
 
 The algorithm for reordering a sequence of marks is:
@@ -373,11 +373,11 @@ The algorithm for reordering a sequence of marks is:
        characters. The subsequence must be moved as a group.
 
 > Note: Unicode describes this mark-reordering operation, the Arabic
-> Mark Transient Reordering Algorithm (AMTRA), in Technical Report 53,
+> Mark Transient Reordering Algorithm (<abbr>AMTRA</abbr>), in Technical Report 53,
 > which describes it in terms that are distinct from standard,
 > Ccc-based mark reordering.
 >
-> Specifically, AMTRA is designated as an operation performed during
+> Specifically, <abbr>AMTRA</abbr> is designated as an operation performed during
 > text rendering only, which therefore does not impact other
 > Unicode-compliance issues such as allowable input sequences or text
 > encoding.
@@ -519,7 +519,7 @@ Finally, the decomposed mark must be reordered as follows:
     the word.
 	
 
-### 5. Applying the language-form substitution features from GSUB ###
+### 5. Applying the language-form substitution features from <abbr>GSUB</abbr> ###
 
 The language-substitution phase applies mandatory substitution
 features using the rules in the font's <abbr>GSUB</abbr> table. In preparation for
@@ -672,7 +672,7 @@ can be disabled by application-level user interfaces.
 
 
 
-### 6. Applying the typographic-form substitution features from GSUB ###
+### 6. Applying the typographic-form substitution features from <abbr>GSUB</abbr> ###
 
 The typographic-substitution phase applies optional substitution
 features using the rules in the font's <abbr>GSUB</abbr> table.
@@ -727,7 +727,7 @@ of bases and marks with precomposed base-and-mark glyphs.
 > Nevertheless, when the active font uses `mset` substitutions, the
 > shaping engine must deal with the situation gracefully.
 
-### 7. Applying the positioning features from GPOS ###
+### 7. Applying the positioning features from <abbr>GPOS</abbr> ###
 
 The positioning stage adjusts the positions of mark and base
 glyphs.
diff --git a/opentype-shaping-myanmar.md b/opentype-shaping-myanmar.md
index f090d61..64e2de1 100644
--- a/opentype-shaping-myanmar.md
+++ b/opentype-shaping-myanmar.md
@@ -14,9 +14,9 @@ runs in the Myanmar script.
   - [The `<mym2>` shaping model](#the-mym2-shaping-model)
       - [1: Identifying syllables and other sequences](#1-identifying-syllables-and-other-sequences)
       - [2: Initial reordering](#2-initial-reordering)
-      - [3: Applying the basic substitution features from GSUB](#3-applying-the-basic-substitution-features-from-gsub)
-      - [4: Applying all remaining substitution features from GSUB](#4-applying-all-remaining-substitution-features-from-gsub)
-      - [5: Applying remaining positioning features from GPOS](#5-applying-remaining-positioning-features-from-gpos)
+      - [3: Applying the basic substitution features from <abbr>GSUB</abbr>](#3-applying-the-basic-substitution-features-from-gsub)
+      - [4: Applying all remaining substitution features from <abbr>GSUB</abbr>](#4-applying-all-remaining-substitution-features-from-gsub)
+      - [5: Applying remaining positioning features from <abbr>GPOS</abbr>](#5-applying-remaining-positioning-features-from-gpos)
   - [The `<mymr>` shaping model](#the-mymr-shaping-model)
 
 
@@ -308,12 +308,13 @@ where an initial "Ra,Halant" sequence without the zero-width joiner
 otherwise would.
 --->
 
-The no-break space is primarily used to display those codepoints that
-are defined as non-spacing (marks, dependent vowels (matras),
-below-base consonant forms, and post-base consonant forms) in an
-isolated context, as an alternative to displaying them superimposed on
-the dotted-circle placeholder. These sequences will match
-"NBSP,ZWJ,Halant,_Consonant_", "NBSP,_mark_", or "NBSP,_matra_".
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display
+those codepoints that are defined as non-spacing (marks, dependent
+vowels (matras), below-base consonant forms, and post-base consonant
+forms) in an isolated context, as an alternative to displaying them
+superimposed on the dotted-circle placeholder. These sequences will
+match "NBSP,ZWJ,Halant,_Consonant_", "NBSP,_mark_", or
+"NBSP,_matra_".
 
 The zero-width space may be used between words — even though no visual
 word spacing results — in order to indicate word breaks within a text
@@ -806,7 +807,7 @@ tagged with the same positioning tag as the closest subsequent consonant.
 With these steps completed, the syllable can be sorted into the final sort order.
 
 
-### 3: Applying the basic substitution features from GSUB ###
+### 3: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -911,7 +912,7 @@ special forms.
 
 
 
-### 4: Applying all remaining substitution features from GSUB ###
+### 4: Applying all remaining substitution features from <abbr>GSUB</abbr> ###
 
 In this stage, the remaining substitution features from the <abbr>GSUB</abbr> table
 are applied. The order in which these features are applied is not
@@ -963,7 +964,7 @@ application-level user interfaces.
 
 
 
-### 5: Applying remaining positioning features from GPOS ###
+### 5: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied. As with the preceding stage, the order in which these
diff --git a/opentype-shaping-nko.md b/opentype-shaping-nko.md
index 4bab001..33e9d29 100644
--- a/opentype-shaping-nko.md
+++ b/opentype-shaping-nko.md
@@ -18,9 +18,9 @@ implementations share.
       - [2. Compound character composition and decomposition](#2-compound-character-composition-and-decomposition)
       - [3. Computing letter joining states](#3-computing-letter-joining-states)
       - [4. Applying the `stch` feature](#4-applying-the-stch-feature)
-      - [5. Applying the language-form substitution features from GSUB](#5-applying-the-language-form-substitution-features-from-gsub)
-      - [6. Applying the typographic-form substitution features from GSUB](#6-applying-the-typographic-form-substitution-features-from-gsub)
-      - [7. Applying the positioning features from GPOS](#7-applying-the-positioning-features-from-gpos)
+      - [5. Applying the language-form substitution features from <abbr>GSUB</abbr>](#5-applying-the-language-form-substitution-features-from-gsub)
+      - [6. Applying the typographic-form substitution features from <abbr>GSUB</abbr>](#6-applying-the-typographic-form-substitution-features-from-gsub)
+      - [7. Applying the positioning features from <abbr>GPOS</abbr>](#7-applying-the-positioning-features-from-gpos)
   
 
 
@@ -212,17 +212,17 @@ This requirement covers:
 
 
 
-The combining grapheme joiner (CGJ) is primarily used to alter the
+The combining grapheme joiner (<abbr>CGJ</abbr>) is primarily used to alter the
 order in which adjacent marks are positioned during the
 mark-reordering stage, in order to adhere to the needs of a
 non-default language orthography.
 
 By default, OpenType shaping reorders sequences of adjacent marks by
 sorting the sequence on the marks' Canonical_Combining_Class (Ccc)
-values. The presence of a CGJ character within a sequence of marks has
+values. The presence of a <abbr>CGJ</abbr> character within a sequence of marks has
 the effect of splitting the sequence into two sequences of marks and,
 therefore, halting any mark-reordering that would have occurred
-between the marks on either side of the CGJ.
+between the marks on either side of the <abbr>CGJ</abbr>.
 
 The zero-width joiner (<abbr>ZWJ</abbr>) is primarily used to force the usage of the
 cursive connecting form of a letter even when the context of the
@@ -260,10 +260,10 @@ when matching sequences of codepoints against the backtrack and
 lookahead sequences of a font's <abbr>GSUB</abbr> or <abbr>GPOS</abbr> lookups.
 
 
-The right-to-left mark (RLM) and left-to-right mark (LRM) are used by
+The right-to-left mark (<abbr>RLM</abbr>) and left-to-right mark (<abbr>LRM</abbr>) are used by
 the Unicode bidirectionality algorithm (BiDi) to indicate the points
 in a text run at which the writing direction changes. Generally
-speaking RLM and LRM codepoints do not interact with shaping.
+speaking <abbr>RLM</abbr> and <abbr>LRM</abbr> codepoints do not interact with shaping.
 
 The no-break space is primarily used to display those codepoints that
 are defined as non-spacing (such as vowel or diacritical marks and "Hamza") in an
@@ -288,7 +288,7 @@ Processing a run of `<nko >` text involves seven top-level stages:
 
 <!--- http://www.unicode.org/reports/tr53/tr53-1.pdf --->
 > Note: because N'Ko does not feature the "Shadda" mark or any
-> marks that belong to _Modifier Combining Marks_ (MCM) classes, this
+> marks that belong to _Modifier Combining Marks_ (<abbr>MCM</abbr>) classes, this
 > stage should not involve any additional work when processing
 > `<nko >` text runs. It is included here to maintain consistency with
 > other scripts that utilize the general Arabic-based shaping model.
@@ -305,7 +305,7 @@ cross-script mark-reordering performed during Unicode
 normalization. The standard Unicode mark-reordering algorithm is based
 on comparing the _Canonical_Combining_Class_ (Ccc) properties of mark
 codepoints, whereas this script-specific reordering utilizes the
-_Modifier_Combining_Mark_ (`MCM`) subclasses specified in the
+_Modifier_Combining_Mark_ (<abbr>MCM</abbr>) subclasses specified in the
 character tables.
 
 The algorithm for reordering a sequence of marks is:
@@ -324,11 +324,11 @@ The algorithm for reordering a sequence of marks is:
        characters. The subsequence must be moved as a group.
 
 > Note: Unicode describes this mark-reordering operation, the Arabic
-> Mark Transient Reordering Algorithm (AMTRA), in Technical Report 53,
+> Mark Transient Reordering Algorithm (<abbr>AMTRA</abbr>), in Technical Report 53,
 > which describes it in terms that are distinct from standard,
 > Ccc-based mark reordering.
 >
-> Specifically, AMTRA is designated as an operation performed during
+> Specifically, <abbr>AMTRA</abbr> is designated as an operation performed during
 > text rendering only, which therefore does not impact other
 > Unicode-compliance issues such as allowable input sequences or text
 > encoding.
@@ -475,7 +475,7 @@ Finally, the decomposed mark must be reordered as follows:
     the word.
 	
 
-### 5. Applying the language-form substitution features from GSUB ###
+### 5. Applying the language-form substitution features from <abbr>GSUB</abbr> ###
 
 The language-substitution phase applies mandatory substitution
 features using the rules in the font's <abbr>GSUB</abbr> table. In preparation for
@@ -593,7 +593,7 @@ can be disabled by application-level user interfaces.
 
 
 
-### 6. Applying the typographic-form substitution features from GSUB ###
+### 6. Applying the typographic-form substitution features from <abbr>GSUB</abbr> ###
 
 The typographic-substitution phase applies optional substitution
 features using the rules in the font's <abbr>GSUB</abbr> table.
@@ -635,7 +635,7 @@ This feature is not used in `<nko >` text.
 This feature is not used in `<nko >` text.
 
 
-### 7. Applying the positioning features from GPOS ###
+### 7. Applying the positioning features from <abbr>GPOS</abbr> ###
 
 The positioning stage adjusts the positions of mark and base
 glyphs.
diff --git a/opentype-shaping-normalization.md b/opentype-shaping-normalization.md
index c237728..1df4465 100644
--- a/opentype-shaping-normalization.md
+++ b/opentype-shaping-normalization.md
@@ -397,7 +397,7 @@ HARFBUZZ logic here: https://github.com/harfbuzz/harfbuzz/src/hb-ot-shape-normal
 
 
 
-#### 4. Normalization-related GSUB features and other font-specific considerations ####
+#### 4. Normalization-related <abbr>GSUB</abbr> features and other font-specific considerations ####
 
 After the decomposition, mark-reordering, and selective
 recomposition stages, OpenType shaping normalization also takes
diff --git a/opentype-shaping-oriya.md b/opentype-shaping-oriya.md
index c50f73a..e40f14d 100644
--- a/opentype-shaping-oriya.md
+++ b/opentype-shaping-oriya.md
@@ -14,10 +14,10 @@ runs in the Oriya script.
   - [The `<ory2>` shaping model](#the-ory2-shaping-model)
       - [1: Identifying syllables and other sequences](#1-identifying-syllables-and-other-sequences)
       - [2: Initial reordering](#2-initial-reordering)
-      - [3: Applying the basic substitution features from GSUB](#3-applying-the-basic-substitution-features-from-gsub)
+      - [3: Applying the basic substitution features from <abbr>GSUB</abbr>](#3-applying-the-basic-substitution-features-from-gsub)
       - [4: Final reordering](#4-final-reordering)
-      - [5: Applying all remaining substitution features from GSUB](#5-applying-all-remaining-substitution-features-from-gsub)
-      - [6: Applying remaining positioning features from GPOS](#6-applying-remaining-positioning-features-from-gpos)
+      - [5: Applying all remaining substitution features from <abbr>GSUB</abbr>](#5-applying-all-remaining-substitution-features-from-gsub)
+      - [6: Applying remaining positioning features from <abbr>GPOS</abbr>](#6-applying-remaining-positioning-features-from-gpos)
   - [The `<orya>` shaping model](#the-orya-shaping-model)
       - [Distinctions from `<ory2>`](#distinctions-from-ory2)
       - [Advice for handling fonts with `<orya>` features only](#advice-for-handling-fonts-with-orya-features-only)
@@ -312,7 +312,7 @@ For example:
     should still be applied if the dependent-vowel codepoint is preceded
     by "Ka,Halant,ZWJ,Tta" in the text run.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
@@ -1192,7 +1192,7 @@ in <abbr>GSUB</abbr> and <abbr>GPOS</abbr> application stages that follow.
 
 
 
-### 3: Applying the basic substitution features from GSUB ###
+### 3: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -1560,7 +1560,7 @@ involve no work when processing `<ory2>` text. It is included here in
 order to maintain compatibility with the other Indic scripts.
 
 
-### 5: Applying all remaining substitution features from GSUB ###
+### 5: Applying all remaining substitution features from <abbr>GSUB</abbr> ###
 
 In this stage, the remaining substitution features from the <abbr>GSUB</abbr> table
 are applied. In preparation for this stage, glyph sequences should be
@@ -1625,7 +1625,7 @@ typographically problematic.
 ![Halant form substitution](/images/oriya/oriya-haln.png)
 
 
-### 6: Applying remaining positioning features from GPOS ###
+### 6: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied.
diff --git a/opentype-shaping-sinhala.md b/opentype-shaping-sinhala.md
index 2c79b52..d596602 100644
--- a/opentype-shaping-sinhala.md
+++ b/opentype-shaping-sinhala.md
@@ -14,10 +14,10 @@ runs in the Sinhala script.
   - [The `<sinh>` shaping model](#the-sinh-shaping-model)
       - [1: Identifying syllables and other sequences](#1-identifying-syllables-and-other-sequences)
       - [2: Initial reordering](#2-initial-reordering)
-      - [3: Applying the basic substitution features from GSUB](#3-applying-the-basic-substitution-features-from-gsub)
+      - [3: Applying the basic substitution features from <abbr>GSUB</abbr>](#3-applying-the-basic-substitution-features-from-gsub)
       - [4: Final reordering](#4-final-reordering)
-      - [5: Applying all remaining substitution features from GSUB](#5-applying-all-remaining-substitution-features-from-gsub)
-      - [6: Applying remaining positioning features from GPOS](#6-applying-remaining-positioning-features-from-gpos)
+      - [5: Applying all remaining substitution features from <abbr>GSUB</abbr>](#5-applying-all-remaining-substitution-features-from-gsub)
+      - [6: Applying remaining positioning features from <abbr>GPOS</abbr>](#6-applying-remaining-positioning-features-from-gpos)
 
 ## General information ##
 
@@ -302,7 +302,7 @@ For example:
     should still be applied if the dependent-vowel codepoint is preceded
     by "Ka,Halant,ZWJ,Tta" in the text run.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
@@ -1010,7 +1010,7 @@ in <abbr>GSUB</abbr> and <abbr>GPOS</abbr> application stages that follow.
 
 
 
-### 3: Applying the basic substitution features from GSUB ###
+### 3: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -1286,7 +1286,7 @@ involve no work when processing `<sinh>` text. It is included here in
 order to maintain compatibility with the other Indic scripts.
 
    
-### 5: Applying all remaining substitution features from GSUB ###
+### 5: Applying all remaining substitution features from <abbr>GSUB</abbr> ###
 
 In this stage, the remaining substitution features from the <abbr>GSUB</abbr> table
 are applied. In preparation for this stage, glyph sequences should be
@@ -1340,7 +1340,7 @@ The `haln` feature is not used in Sinhala.
 > and may be disabled in the application by user preference.
 
 
-### 6: Applying remaining positioning features from GPOS ###
+### 6: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied.
diff --git a/opentype-shaping-syriac.md b/opentype-shaping-syriac.md
index 86c525b..fe44196 100644
--- a/opentype-shaping-syriac.md
+++ b/opentype-shaping-syriac.md
@@ -18,9 +18,9 @@ implementations share.
       - [2. Compound character composition and decomposition](#2-compound-character-composition-and-decomposition)
       - [3. Computing letter joining states](#3-computing-letter-joining-states)
       - [4. Applying the `stch` feature](#4-applying-the-stch-feature)
-      - [5. Applying the language-form substitution features from GSUB](#5-applying-the-language-form-substitution-features-from-gsub)
-      - [6. Applying the typographic-form substitution features from GSUB](#6-applying-the-typographic-form-substitution-features-from-gsub)
-      - [7. Applying the positioning features from GPOS](#7-applying-the-positioning-features-from-gpos)
+      - [5. Applying the language-form substitution features from <abbr>GSUB</abbr>](#5-applying-the-language-form-substitution-features-from-gsub)
+      - [6. Applying the typographic-form substitution features from <abbr>GSUB</abbr>](#6-applying-the-typographic-form-substitution-features-from-gsub)
+      - [7. Applying the positioning features from <abbr>GPOS</abbr>](#7-applying-the-positioning-features-from-gpos)
   
 
 
@@ -279,17 +279,17 @@ Modern texts may also make use of Arabic punctuation marks, and texts
 using Syriac to write Arabic (called "Garshuni") may also employ
 Arabic ḥarakah (vowel) marks.
 
-The combining grapheme joiner (CGJ) is primarily used to alter the
+The combining grapheme joiner (<abbr>CGJ</abbr>) is primarily used to alter the
 order in which adjacent marks are positioned during the
 mark-reordering stage, in order to adhere to the needs of a
 non-default language orthography.
 
 By default, OpenType shaping reorders sequences of adjacent marks by
 sorting the sequence on the marks' Canonical_Combining_Class (Ccc)
-values. The presence of a CGJ character within a sequence of marks has
+values. The presence of a <abbr>CGJ</abbr> character within a sequence of marks has
 the effect of splitting the sequence into two sequences of marks and,
 therefore, halting any mark-reordering that would have occurred
-between the marks on either side of the CGJ.
+between the marks on either side of the <abbr>CGJ</abbr>.
 
 The zero-width joiner (<abbr>ZWJ</abbr>) is primarily used to force the usage of the
 cursive connecting form of a letter even when the context of the
@@ -327,10 +327,10 @@ when matching sequences of codepoints against the backtrack and
 lookahead sequences of a font's <abbr>GSUB</abbr> or <abbr>GPOS</abbr> lookups.
 
 
-The right-to-left mark (RLM) and left-to-right mark (LRM) are used by
+The right-to-left mark (<abbr>RLM</abbr>) and left-to-right mark (<abbr>LRM</abbr>) are used by
 the Unicode bidirectionality algorithm (BiDi) to indicate the points
 in a text run at which the writing direction changes. Generally
-speaking RLM and LRM codepoints do not interact with shaping.
+speaking <abbr>RLM</abbr> and <abbr>LRM</abbr> codepoints do not interact with shaping.
 
 The no-break space is primarily used to display those codepoints that
 are defined as non-spacing (such as vowel points or diacritical marks) in an
@@ -373,7 +373,7 @@ cross-script mark-reordering performed during Unicode
 normalization. The standard Unicode mark-reordering algorithm is based
 on comparing the _Canonical_Combining_Class_ (Ccc) properties of mark
 codepoints, whereas this script-specific reordering utilizes the
-_Modifier_Combining_Mark_ (`MCM`) subclasses specified in the
+_Modifier_Combining_Mark_ (<abbr>MCM</abbr>) subclasses specified in the
 character tables.
 
 The algorithm for reordering a sequence of marks is:
@@ -392,11 +392,11 @@ The algorithm for reordering a sequence of marks is:
        characters. The subsequence must be moved as a group.
 
 > Note: Unicode describes this mark-reordering operation, the Arabic
-> Mark Transient Reordering Algorithm (AMTRA), in Technical Report 53,
+> Mark Transient Reordering Algorithm (<abbr>AMTRA</abbr>), in Technical Report 53,
 > which describes it in terms that are distinct from standard,
 > Ccc-based mark reordering.
 >
-> Specifically, AMTRA is designated as an operation performed during
+> Specifically, <abbr>AMTRA</abbr> is designated as an operation performed during
 > text rendering only, which therefore does not impact other
 > Unicode-compliance issues such as allowable input sequences or text
 > encoding.
@@ -533,7 +533,7 @@ Finally, the decomposed mark must be reordered as follows:
     the word.
 	
 
-### 5. Applying the language-form substitution features from GSUB ###
+### 5. Applying the language-form substitution features from <abbr>GSUB</abbr> ###
 
 The language-substitution phase applies mandatory substitution
 features using the rules in the font's <abbr>GSUB</abbr> table. In preparation for
@@ -664,7 +664,7 @@ can be disabled by application-level user interfaces.
 
 
 
-### 6. Applying the typographic-form substitution features from GSUB ###
+### 6. Applying the typographic-form substitution features from <abbr>GSUB</abbr> ###
 
 The typographic-substitution phase applies optional substitution
 features using the rules in the font's <abbr>GSUB</abbr> table.
@@ -705,7 +705,7 @@ This feature is not used in `<syrc>` text.
 This feature is not used in `<syrc>` text.
 
 
-### 7. Applying the positioning features from GPOS ###
+### 7. Applying the positioning features from <abbr>GPOS</abbr> ###
 
 The positioning stage adjusts the positions of mark and base
 glyphs.
diff --git a/opentype-shaping-tamil.md b/opentype-shaping-tamil.md
index 0fc0502..af9f536 100644
--- a/opentype-shaping-tamil.md
+++ b/opentype-shaping-tamil.md
@@ -14,10 +14,10 @@ runs in the Tamil script.
   - [The `<tml2>` shaping model](#the-tml2-shaping-model)
       - [1: Identifying syllables and other sequences](#1-identifying-syllables-and-other-sequences)
       - [2: Initial reordering](#2-initial-reordering)
-      - [3: Applying the basic substitution features from GSUB](#3-applying-the-basic-substitution-features-from-gsub)
+      - [3: Applying the basic substitution features from <abbr>GSUB</abbr>](#3-applying-the-basic-substitution-features-from-gsub)
       - [4: Final reordering](#4-final-reordering)
-      - [5: Applying all remaining substitution features from GSUB](#5-applying-all-remaining-substitution-features-from-gsub)
-      - [6: Applying remaining positioning features from GPOS](#6-applying-remaining-positioning-features-from-gpos)
+      - [5: Applying all remaining substitution features from <abbr>GSUB</abbr>](#5-applying-all-remaining-substitution-features-from-gsub)
+      - [6: Applying remaining positioning features from <abbr>GPOS</abbr>](#6-applying-remaining-positioning-features-from-gpos)
   - [The `<taml>` shaping model](#the-taml-shaping-model)
       - [Distinctions from `<tml2>`](#distinctions-from-tml2)
       - [Advice for handling fonts with `<taml>` features only](#advice-for-handling-fonts-with-taml-features-only)
@@ -331,7 +331,7 @@ For example:
     should still be applied if the dependent-vowel codepoint is preceded
     by "Ka,Halant,ZWJ,Tta" in the text run.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
@@ -1115,7 +1115,7 @@ in <abbr>GSUB</abbr> and <abbr>GPOS</abbr> application stages that follow.
 
 
 
-### 3: Applying the basic substitution features from GSUB ###
+### 3: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -1503,7 +1503,7 @@ involve no work when processing `<tml2>` text. It is included here in
 order to maintain compatibility with the other Indic scripts.
 
 
-### 5: Applying all remaining substitution features from GSUB ###
+### 5: Applying all remaining substitution features from <abbr>GSUB</abbr> ###
 
 In this stage, the remaining substitution features from the <abbr>GSUB</abbr> table
 are applied. In preparation for this stage, glyph sequences should be
@@ -1562,7 +1562,7 @@ typographically problematic.
 > point. However, `calt` is not mandatory for correct Tamil shaping
 > and may be disabled in the application by user preference.
 
-### 6: Applying remaining positioning features from GPOS ###
+### 6: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied.
diff --git a/opentype-shaping-telugu.md b/opentype-shaping-telugu.md
index a7db650..bf02732 100644
--- a/opentype-shaping-telugu.md
+++ b/opentype-shaping-telugu.md
@@ -14,10 +14,10 @@ runs in the Telugu script.
   - [The `<tel2>` shaping model](#the-tel2-shaping-model)
       - [1: Identifying syllables and other sequences](#1-identifying-syllables-and-other-sequences)
       - [2: Initial reordering](#2-initial-reordering)
-      - [3: Applying the basic substitution features from GSUB](#3-applying-the-basic-substitution-features-from-gsub)
+      - [3: Applying the basic substitution features from <abbr>GSUB</abbr>](#3-applying-the-basic-substitution-features-from-gsub)
       - [4: Final reordering](#4-final-reordering)
-      - [5: Applying all remaining substitution features from GSUB](#5-applying-all-remaining-substitution-features-from-gsub)
-      - [6: Applying remaining positioning features from GPOS](#6-applying-remaining-positioning-features-from-gpos)
+      - [5: Applying all remaining substitution features from <abbr>GSUB</abbr>](#5-applying-all-remaining-substitution-features-from-gsub)
+      - [6: Applying remaining positioning features from <abbr>GPOS</abbr>](#6-applying-remaining-positioning-features-from-gpos)
   - [The `<telu>` shaping model](#the-telu-shaping-model)
       - [Distinctions from `<tel2>`](#distinctions-from-tel2)
       - [Advice for handling fonts with `<telu>` features only](#advice-for-handling-fonts-with-telu-features-only)
@@ -325,7 +325,7 @@ For example:
     should still be applied if the dependent-vowel codepoint is preceded
     by "Ka,Halant,ZWJ,Tta" in the text run.
 
-The no-break space (NBSP) is primarily used to display those
+The no-break space (<abbr>NBSP</abbr>) is primarily used to display those
 codepoints that are defined as non-spacing (marks, dependent vowels
 (matras), below-base consonant forms, and post-base consonant forms)
 in an isolated context, as an alternative to displaying them
@@ -1136,7 +1136,7 @@ in <abbr>GSUB</abbr> and <abbr>GPOS</abbr> application stages that follow.
 
 
 
-### 3: Applying the basic substitution features from GSUB ###
+### 3: Applying the basic substitution features from <abbr>GSUB</abbr> ###
 
 The basic-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -1494,7 +1494,7 @@ involve no work when processing `<tel2>` text. It is included here in
 order to maintain compatibility with the other Indic scripts.
 
 
-### 5: Applying all remaining substitution features from GSUB ###
+### 5: Applying all remaining substitution features from <abbr>GSUB</abbr> ###
 
 In this stage, the remaining substitution features from the <abbr>GSUB</abbr> table
 are applied. In preparation for this stage, glyph sequences should be
@@ -1553,7 +1553,7 @@ typographically problematic.
 
 ![Halant form ligation](/images/telugu/telugu-haln.png)
 
-### 6: Applying remaining positioning features from GPOS ###
+### 6: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied.
diff --git a/opentype-shaping-thai-lao.md b/opentype-shaping-thai-lao.md
index 8170737..16b3b0b 100644
--- a/opentype-shaping-thai-lao.md
+++ b/opentype-shaping-thai-lao.md
@@ -11,18 +11,18 @@ runs in the Thai and Lao scripts.
   - [Glyph classification](#glyph-classification)
       - [Shaping classes and subclasses](#shaping-classes-and-subclasses)
       - [Mark combining classes](#mark-combining-classes)
-      - [PUA fallback classifications](#pua-fallback-classifications)
+      - [<abbr>PUA</abbr> fallback classifications](#pua-fallback-classifications)
       - [Thai and Lao character tables](#thai-and-lao-character-tables)
   - [The `<thai>`/`<lao >` shaping model](#the-thailao-shaping-model)
-      - [1: Applying the language substitution features from GSUB](#1-applying-the-language-substitution-features-from-gsub)
+      - [1: Applying the language substitution features from <abbr>GSUB</abbr>](#1-applying-the-language-substitution-features-from-gsub)
       - [2: Decomposing all Am vowel signs](#2-decomposing-all-am-vowel-signs)
       - [3: Reordering sequences of marks](#3-reordering-sequences-of-marks)
-      - [4: Applying all positioning features from GPOS](#4-applying-all-positioning-features-from-gpos)
-  - [The PUA fallback shaping model](#the-pua-fallback-shaping-model)
+      - [4: Applying all positioning features from <abbr>GPOS</abbr>](#4-applying-all-positioning-features-from-gpos)
+  - [The <abbr>PUA</abbr> fallback shaping model](#the-pua-fallback-shaping-model)
       - [Contextual replacement rules](#contextual-replacement-rules)
 	  - [1. Decomposing all Am vowel signs](#1-decomposing-all-am-vowel-signs)
       - [2. Reordering sequences of marks](#2-reordering-sequences-of-marks)
-      - [3. Remapping codepoints to the appropriate PUA alternates](#3-remapping-codepoints-to-the-appropriate-pua-alternates)
+      - [3. Remapping codepoints to the appropriate <abbr>PUA</abbr> alternates](#3-remapping-codepoints-to-the-appropriate-pua-alternates)
 
 
 
@@ -56,8 +56,8 @@ space.
 
 A significant number of  older Thai fonts that do not use the OpenType
 shaping model are still in usage; these fonts employ the Unicode
-"Private Use Area" (`PUA`) to store contextual forms of
-characters. Shaping engines may implement this PUA-base shaping model
+"Private Use Area" (<abbr>PUA</abbr>) to store contextual forms of
+characters. Shaping engines may implement this <abbr>PUA</abbr>-base shaping model
 as a fallback mechanism when such fonts are encountered.
 
 
@@ -220,9 +220,9 @@ This ensures that "Sara U" or "Sara Uu" codepoints adjacent to
 > handle the Phinthu-reordering issue in a different manner.
 
 
-### PUA fallback classifications ###
+### <abbr>PUA</abbr> fallback classifications ###
 
-Older Thai fonts that implement the PUA-substitution fallback method
+Older Thai fonts that implement the <abbr>PUA</abbr>-substitution fallback method
 rather than modern OpenType script shaping rules incorporate
 subclasses for consonants that indicate whether or not the consonant
 includes an ascender, a normal descender, or a removable descender.
@@ -261,7 +261,7 @@ and `<lao >` text runs:
 
 The tables list each codepoint along with its Unicode general
 category, its shaping class, its mark-placement subclass, and its
-PUA-fallback category. The codepoint's Unicode name and an example
+<abbr>PUA</abbr>-fallback category. The codepoint's Unicode name and an example
 glyph are also provided.
 
 For example:
@@ -292,11 +292,11 @@ category_ column are categorized as non-spacing; marks tagged with
 [Mc] are categorized as spacing-combining.
 
 The _PUA_ column indicates which, if any, fallback-shaping category
-the codepoint belongs to when found in older fonts using the PUA
-fallback shaping scheme. Note that the PUA method was employed only
-for Thai fonts, so Lao codepoints do not have a PUA fallback-shaping
+the codepoint belongs to when found in older fonts using the <abbr>PUA</abbr>
+fallback shaping scheme. Note that the <abbr>PUA</abbr> method was employed only
+for Thai fonts, so Lao codepoints do not have a <abbr>PUA</abbr> fallback-shaping
 category. Thai codepoints with a _null_ in the _PUA_ column were not
-used in the PUA fallback-shaping scheme and evoke no special behavior
+used in the <abbr>PUA</abbr> fallback-shaping scheme and evoke no special behavior
 from the shaping engine.
 
 Some codepoints in the tables use a _Shaping class_ that
@@ -365,7 +365,7 @@ In addition to valid syllables, standalone sequences may occur, such
 as when an isolated codepoint is shown in example text.
 
 
-### 1: Applying the language substitution features from GSUB ###
+### 1: Applying the language substitution features from <abbr>GSUB</abbr> ###
 
 The language-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -484,7 +484,7 @@ move the
 --->
 
 
-### 4: Applying all positioning features from GPOS ###
+### 4: Applying all positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied. As with the preceding stage, the order in which these
@@ -518,17 +518,17 @@ same base glyph.
 
 
 
-## The PUA fallback shaping model ##
+## The <abbr>PUA</abbr> fallback shaping model ##
 
 A significant number of  older Thai fonts that do not use the OpenType
 shaping model are still in usage; these fonts employ the Unicode
-"Private Use Area" (`PUA`) to store contextual forms of
+"Private Use Area" (<abbr>PUA</abbr>) to store contextual forms of
 characters.
 
-The PUA shaping model is described at
+The <abbr>PUA</abbr> shaping model is described at
 [linux.thai.net/~thep/th-otf/shaping.html](https://linux.thai.net/~thep/th-otf/shaping.html)
 . It relies on a set of pre-determined mappings from the codepoints in the
-Unicode Thai block to codepoints in the PUA.
+Unicode Thai block to codepoints in the <abbr>PUA</abbr>.
 
 For consonants, these alternate-glyph mappings depend on whether or
 not the consonant includes an ascender, a normal descender, or a
@@ -562,7 +562,7 @@ table](character-tables/character-tables-thai.md#thai-character-table).
 ## Contextual replacement rules ##
 
 Codepoints in the Thai Block can be mapped to one of several alternate
-PUA codepoints depending on context:
+<abbr>PUA</abbr> codepoints depending on context:
 
   - A tone marker that does not follow an above-base vowel sign may be
     mapped to an alternate that is positioned lower, closer to the top
@@ -603,7 +603,7 @@ These replacements take the place of both <abbr>GSUB</abbr> substitutions and <a
 positioning in modern OpenType fonts.
 
 Shaping engines can replace the original codepoints with the
-appropriate alternates from the PUA block by testing for the above
+appropriate alternates from the <abbr>PUA</abbr> block by testing for the above
 conditions. 
 
 With each consonant, vowel, and mark character correctly classified,
@@ -613,7 +613,7 @@ There are three top-level stages:
 
 1. Decomposing all Am vowel signs
 2. Reordering sequences of marks
-3. Remapping codepoints to the appropriate PUA alternates
+3. Remapping codepoints to the appropriate <abbr>PUA</abbr> alternates
 
 
 ### 1. Decomposing all Am vowel signs ###
@@ -654,7 +654,7 @@ for possible reordering.
 > the text run, or by performing a sorting step at this stage.
 
 
-### 3. Remapping codepoints to the appropriate PUA alternates ###
+### 3. Remapping codepoints to the appropriate <abbr>PUA</abbr> alternates ###
 
 The contextual replacement rules described above can be implemented in
 a pair of state machines, one for above-base replacement moves and one
@@ -722,7 +722,7 @@ follow a consonant:
 
 
 When the necessary replacement action for each codepoint has been
-determined, codepoints can be replaced with the PUA codepoints from
+determined, codepoints can be replaced with the <abbr>PUA</abbr> codepoints from
 the following table.
 
 Note that Windows fonts and MacOS fonts used different mappings.
diff --git a/opentype-shaping-tibetan.md b/opentype-shaping-tibetan.md
index 53f42be..c5f608c 100644
--- a/opentype-shaping-tibetan.md
+++ b/opentype-shaping-tibetan.md
@@ -12,9 +12,9 @@ runs in the Tibetan script.
       - [Shaping classes and subclasses](#shaping-classes-and-subclasses)
       - [Tibetan character tables](#tibetan-character-tables)
   - [The `<tibt>` shaping model](#the-tibt-shaping-model)
-      - [1: Applying the language substitution features from GSUB](#1-applying-the-language-substitution-features-from-gsub)
-      - [2: Applying all basic substitution features from GSUB](#2-applying-all-basic-substitution-features-from-gsub)
-      - [3: Applying remaining positioning features from GPOS](#3-applying-remaining-positioning-features-from-gpos)
+      - [1: Applying the language substitution features from <abbr>GSUB</abbr>](#1-applying-the-language-substitution-features-from-gsub)
+      - [2: Applying all basic substitution features from <abbr>GSUB</abbr>](#2-applying-all-basic-substitution-features-from-gsub)
+      - [3: Applying remaining positioning features from <abbr>GPOS</abbr>](#3-applying-remaining-positioning-features-from-gpos)
 
 
 ## General information ##
@@ -360,7 +360,7 @@ as when an isolated codepoint is shown in example text.
 
 
 
-### 1: Applying the language substitution features from GSUB ###
+### 1: Applying the language substitution features from <abbr>GSUB</abbr> ###
 
 The language-substitution stage applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -417,7 +417,7 @@ glyphs.
 
 
 
-### 2: Applying all basic substitution features from GSUB ###
+### 2: Applying all basic substitution features from <abbr>GSUB</abbr> ###
 
 In this stage, the basic substitution features from the <abbr>GSUB</abbr> table
 are applied. The order in which these features are applied is not
@@ -466,7 +466,7 @@ application-level user interfaces.
 
 
 
-### 3: Applying remaining positioning features from GPOS ###
+### 3: Applying remaining positioning features from <abbr>GPOS</abbr> ###
 
 In this stage, mark positioning, kerning, and other <abbr>GPOS</abbr> features are
 applied. As with the preceding stage, the order in which these
diff --git a/opentype-shaping-use.md b/opentype-shaping-use.md
index f62303f..1808687 100644
--- a/opentype-shaping-use.md
+++ b/opentype-shaping-use.md
@@ -10,20 +10,20 @@ model.
   - [General information](#general-information)
   - [Terminology](#terminology)
   - [Glyph classification](#glyph-classification)
-  - [The USE shaping model](#the-use-shaping-model)
+  - [The <abbr>USE</abbr> shaping model](#the-use-shaping-model)
       - [1: Split vowel decomposition](#1-split-vowel-decomposition)
       - [2: Cluster identification](#2-cluster-identification)
       - [3: Basic cluster formation](#3-basic-cluster-formation)
-	      - [3.1: Applying the basic pre-processing features from GSUB](#31-applying-the-basic-pre-processing-features-from-gsub)
-          - [3.2: Applying the basic reordering features from GSUB](#32-applying-the-basic-reordering-features-from-gsub)
-          - [3.3: Applying the basic orthographic features from GSUB](#33-applying-the-basic-orthographic-features-from-gsub)
+	      - [3.1: Applying the basic pre-processing features from <abbr>GSUB</abbr>](#31-applying-the-basic-pre-processing-features-from-gsub)
+          - [3.2: Applying the basic reordering features from <abbr>GSUB</abbr>](#32-applying-the-basic-reordering-features-from-gsub)
+          - [3.3: Applying the basic orthographic features from <abbr>GSUB</abbr>](#33-applying-the-basic-orthographic-features-from-gsub)
 	  - [4: Glyph reordering](#4-glyph-reordering)
-	      - [4.1: Applying the reordering features from GSUB](#41-applying-the-reordering-features-from-gsub)
+	      - [4.1: Applying the reordering features from <abbr>GSUB</abbr>](#41-applying-the-reordering-features-from-gsub)
 	      - [4.2: Performing property-based reordering moves](#42-performing-property-based-reordering-moves)
 	  - [5: Final feature application](#5-final-feature-application)
-	      - [5.1: Applying the final topographic features from GSUB](#51-applying-the-final-topographic-features-from-gsub)
-	      - [5.2: Applying the final typographic-presentation features from GSUB](#52-applying-the-final-typographic-presentation-features-from-gsub)
-	      - [5.3: Applying the final positioning features from GPOS](#53-applying-the-final-positioning-features-from-gpos)
+	      - [5.1: Applying the final topographic features from <abbr>GSUB</abbr>](#51-applying-the-final-topographic-features-from-gsub)
+	      - [5.2: Applying the final typographic-presentation features from <abbr>GSUB</abbr>](#52-applying-the-final-typographic-presentation-features-from-gsub)
+	      - [5.3: Applying the final positioning features from <abbr>GPOS</abbr>](#53-applying-the-final-positioning-features-from-gpos)
   
   
   
@@ -120,7 +120,7 @@ In addition, the Unicode Character Decomposition Mapping (<abbr>UCDM</abbr>) is
 all split vowels.
 
 
-### USE overrides ###
+### <abbr>USE</abbr> overrides ###
 
 Although, in general, the <abbr>USE</abbr> shaping model relies on the <abbr>UGC</abbr>, <abbr>UISC</abbr>,
 and <abbr>UIPC</abbr> properties, the <abbr>USE</abbr> model makes a small set of standardized
@@ -159,7 +159,7 @@ correct results.
 | | | | | | |
 
 
-### USE classification table ###
+### <abbr>USE</abbr> classification table ###
 
 The following table lists the classes utilized in the <abbr>USE</abbr> shaping
 model, along with a definition for each class. The class definitions
@@ -170,7 +170,7 @@ The symbols given in the "Symbol" column for each class may be used to
 express cluster-matching rules or other algorithms.
 
 Vowels and modifiers may be further subclassified as described in the
-[USE subclasses table](#use-subclasses-table) below.
+[<abbr>USE</abbr> subclasses table](#use-subclasses-table) below.
 
 
 | USE classification        | Symbol | Definition                                                                                                    |
@@ -202,7 +202,7 @@ Vowels and modifiers may be further subclassified as described in the
 | | | |
 
 
-### USE subclasses table ###
+### <abbr>USE</abbr> subclasses table ###
 
 Vowels and modifiers may be further subclassified based on their
 position relative to base characters. The subclasses incorporated in
@@ -244,7 +244,7 @@ vowel must be decomposed into its components.
 
 
 
-## The USE shaping model ##
+## The <abbr>USE</abbr> shaping model ##
 
 The <abbr>USE</abbr> shaping model consists of five top-level stages.
 
@@ -368,7 +368,7 @@ shaping stages occurs on a per-cluster basis.
 The basic cluster formation stage is used to apply fundamental
 substitutions necessary for script and language correctness.
 
-#### 3.1: Applying the basic pre-processing features from GSUB ####
+#### 3.1: Applying the basic pre-processing features from <abbr>GSUB</abbr> ####
 
 The basic pre-processing step applies mandatory substitution features
 using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this 
@@ -420,7 +420,7 @@ rules designed to match them in subsequences. Therefore, this
 feature must be applied before all other many-to-one substitutions.
 
 
-#### 3.2: Applying the basic reordering features from GSUB ####
+#### 3.2: Applying the basic reordering features from <abbr>GSUB</abbr> ####
 
 The basic reordering step applies mandatory substitution features from
 <abbr>GSUB</abbr> that affect reordering elements.
@@ -454,7 +454,7 @@ forms.
 > stage. 
 
 
-#### 3.3: Applying the basic orthographic features from GSUB ####
+#### 3.3: Applying the basic orthographic features from <abbr>GSUB</abbr> ####
 
 The basic orthographic step applies substitution features using the
 rules in the font's <abbr>GSUB</abbr> table. In preparation for this stage, glyph
@@ -504,7 +504,7 @@ performed in this stage, which is broken into two distinct steps:
 2. Performing property-based reordering moves
 
 
-#### 4.1 Applying the reordering features from GSUB ####
+#### 4.1 Applying the reordering features from <abbr>GSUB</abbr> ####
 
 In this step, the reordering moves corresponding to the
 glyph-reordering features in <abbr>GSUB</abbr> are performed.
@@ -594,7 +594,7 @@ connected scripts, applying typographic-presentation features from
 <abbr>GSUB</abbr>, and applying positioning features from <abbr>GPOS</abbr>.
 
 
-#### 5.1: Applying the final topographic features from GSUB ####
+#### 5.1: Applying the final topographic features from <abbr>GSUB</abbr> ####
 
 For connected scripts, this step applies the substitutions to select
 the correct topographic form for each glyph, based on its position in
@@ -612,7 +612,7 @@ for each codepoint.
 	medi
 	fina
 
-#### 5.2: Applying the final typographic-presentation features from GSUB ####
+#### 5.2: Applying the final typographic-presentation features from <abbr>GSUB</abbr> ####
 
 The final typographic-presentation step applies mandatory substitution
 features using the rules in the font's <abbr>GSUB</abbr> table. In preparation for this
@@ -687,7 +687,7 @@ application-level user interfaces.
 
 
 
-#### 5.3: Applying the final positioning features from GPOS ####
+#### 5.3: Applying the final positioning features from <abbr>GPOS</abbr> ####
 
 	curs
 	dist