Moving from io to tempfile to respect pandoc fileoutput constraint

Author: guzmud

README.md

@@ -1,69 +1,43 @@
 # Django Panpub
 
-**Warning: early-stage dev, no migration files provided**
+**Warning: early-stage dev**
 
-Panpub, from *pan* ("all, of everything") and *publishing*. A Django publishing app providing united streamlined outputs from centralised normalised inputs.
+Panpub, from *pan* ("all, of everything") and *publishing*. A Django publishing app providing united streamlined outputs from centralised normalised inputs, with an embedded system of claims regarding each content and role (*creator*, *curator*, *mediator*).
 
 
 ## Directions
 
 * **Medium openness**: widely-used media - writing, picture, audio and video - with outside link as a fall-back option
+* **Access first**: various commonly-used formats for import and export
 * **Living culture**: claims of participation (whether as creator, curator or mediator) available for each unique content
-* **Batteries included**: delicious base of models, decorators and templatetags, with its standard topping of views, forms, urls and templates
+* **Batteries included**: delicious base of models, decorators and templatetags, with its standard topping of views, forms, urls and templates for reuse
 
 
 ## Achieved features
-* Panpub ecosystem : data and media export
-* Text model : upload to markdown through pandoc
+* Panpub ecosystem: claim system, corpus creation, system-wide exporting
+* Text model: use of pandoc as the middleman, upload and download
 
 
 ## Ecosystem
 
-Models:
-* Corpus
-* Content
-* Crafter
-* Claim
-
-Decorator:
-* has_any_claim
-* has_creator_claim
-* has_curator_claim
-* has_mediator_claim
-
-Templatetags:
-* crafterworks
+* models: Corpus, Content, Crafter, Claim
+* decorators: has_any_claim, has_creator_claim, has_curator_claim, has_mediator_claim
+* templatetags: crafterworks
+* utils: panpub_export
 
 
 ## Writing
 
-Main class:
-* Text
-
-Available inputs:
-* .md (*pandoc-standard* or *github-flavored*)
-* .latex
-* .docx
-* .odt
-* .html
-* .rst
-* native Haskell
-
-In-house format:
-* markdown (*pandoc-standard*)
-
-Available outputs: (*none implemented yet*)
-* md
-* html
-* pdf
-* .ebook
-
-Templatetags:
-* craftertexts
+* main class: Text
+* inputs: .md (*pandoc-standard* or *github-flavored*), .latex, .docx, .odt, .html, .rst, native Haskell
+* in-house: markdown (*pandoc-standard*)
+* ouputs: .md (*pandoc-standard* or *github-flavored*), .html, .docx, .epub, .odt
+* templatetags: craftertexts
 
 
 ## Spoilers
 
+* Text medium: fixing integration of xetex and adding .pdf to the outputs, extending .md and .latex plugins
 * Picture medium: Pillow-fed input, with opt-in EXIF-cleaner and LSB-smoother (a.k.a. *anti-steganography**)
 * Dataset medium: Tablib-fed input
 * pandoc auto-installation

panpub/models.py

@@ -1,6 +1,8 @@
 #!/usr/bin/env python3
 # -*- coding: utf-8 -*-
 
+import pathlib
+import tempfile
 from io import StringIO
 from sys import getsizeof
 
@@ -147,30 +149,34 @@ def filefriendly_name(self):
         return slugify(self.name)
 
     def available_pubformats(self):
-        # docx, epub, odt require outputfile instead of var
         # pdf requires xetex
         return ('gfm',
                 'html',
                 'markdown',
+                'docx',
+                'epub',
+                'odt',
                 )
 
     def export(self, pubformat='markdown'):
         if pubformat not in self.available_pubformats():
             raise Exception
         try:
-            data = pypandoc.convert_file(self.document.path,
-                                         pubformat,
-                                         format='md')
-            datafile = StringIO(data)
+            with tempfile.NamedTemporaryFile() as f:
+                pypandoc.convert_file(self.document.path,
+                                      pubformat,
+                                      format='md',
+                                      outputfile=pathlib.Path(tempfile.tempdir, f.name).as_posix())
+                f.seek(0)
+                datafile = f.read()
         except Exception:
             raise Exception
         else:
-            filelen = len(datafile.getvalue())
+            filelen = len(datafile)
             filename = '{}.{}'.format(self.filefriendly_name(),
                                       pubformat)
-            datafile.seek(0)
-            filedata = datafile.getvalue()
-            return filedata, filename, filelen
+            filedata = datafile
+            return datafile, filename, filelen
 
 
 class Picture(Content):

panpub/urls.py

@@ -1,7 +1,6 @@
 #!/usr/bin/env python3
 # -*- coding: utf-8 -*-
 from django.conf.urls import url
-from django.views.generic import TemplateView
 
 from panpub import views
 

panpub/views.py

@@ -73,6 +73,7 @@ def text_export(request, text_id, export_format):
 
     if not text.ready:
         raise PermissionDenied  # unless user has claim ?
+
     if export_format not in text.available_pubformats():
         raise Http404("Export format requested unavailable.")