Merge branch 'stable'

CHANGES.rst

@@ -4,14 +4,18 @@ Release History
 Release 0.3.0.dev
 ~~~~~~~~~~~~~~~~~
 
-Release 0.2.1.dev
+Release 0.2.2.dev
 ~~~~~~~~~~~~~~~~~
 
+Release 0.2.1 (2016-08-18)
+~~~~~~~~~~~~~~~~~~~~~~~~~~
+
 New Features:
 
 * optionally limit the width of large images and make use of this to simulate
   the Sphinx LaTeX builder behavior (#46)
 * reStructuredText/Sphinx: support for images with hyperlinks (#49)
+* record the styled page numbers in the PDF as page labels (#41)
 * unsupported Python versions: prevent installation where possible (sdist)
   or exit on import (wheel)
 * support Python 3.6
@@ -22,6 +26,7 @@ Bugfixes:
   option can actually be used
 * Fix #47: ClassNotFound exception in Literal_Block.lexer_getter()
 * Fix #45: Images that don't fit are still placed on the page
+* don't warn about duplicate style matches that resolve to the same style
 
 
 Release 0.2.0 (2016-08-10)

src/rinoh/backend/pdf/__init__.py

@@ -8,7 +8,7 @@
 
 import math
 
-from io import StringIO, BytesIO
+from io import BytesIO
 from contextlib import contextmanager
 
 from . import cos
@@ -107,20 +107,37 @@ def _create_outline_level(self, sections_tree, parent, top_level):
         parent['Count'] = cos.Integer(count if top_level else - count)
 
     def write(self, file):
-        for page in self.pages:
+        from ... import number
+        page_number_formats = {number.NUMBER: cos.DECIMAL_ARABIC,
+                               number.CHARACTER_LC: cos.LOWERCASE_LETTERS,
+                               number.CHARACTER_UC: cos.UPPERCASE_LETTERS,
+                               number.ROMAN_LC: cos.LOWERCASE_ROMAN,
+                               number.ROMAN_UC: cos.UPPERCASE_ROMAN}
+
+        page_labels = self.cos_document.catalog['PageLabels']['Nums']
+        last_number_format = None
+        for index, page in enumerate(self.pages):
             contents = cos.Stream(filter=FlateDecode())
             contents.write(page.canvas.getvalue())
             page.cos_page['Contents'] = contents
+            if page.number_format != last_number_format:
+                pdf_number_format = page_number_formats[page.number_format]
+                page_labels.append(cos.Integer(index))
+                page_labels.append(cos.PageLabel(pdf_number_format,
+                                                 start=page.number))
+                last_number_format = page.number_format
         self.cos_document.write(file)
 
 
 class Page(object):
-    def __init__(self, backend_document, width, height):
+    def __init__(self, backend_document, width, height, number, number_format):
         self.backend_document = backend_document
         cos_pages = backend_document.cos_document.catalog['Pages']
         self.cos_page = cos_pages.new_page(float(width), float(height))
         self.width = width
         self.height = height
+        self.number = number
+        self.number_format = number_format
         self.canvas = PageCanvas(self)
         self.backend_document.pages.append(self)
 

src/rinoh/backend/pdf/cos.py

@@ -457,6 +457,8 @@ class Document(dict):
 
     def __init__(self, creator):
         self.catalog = Catalog()
+        self.catalog['PageLabels'] = Dictionary(indirect=True)
+        self.catalog['PageLabels']['Nums'] = Array()
         self.info = Dictionary(indirect=True)
         self.timestamp = time.time()
         self.set_info('Creator', creator)
@@ -605,6 +607,26 @@ def __init__(self, parent, width, height):
                                   Real(width), Real(height)])
 
 
+DECIMAL_ARABIC = Name('D')
+UPPERCASE_ROMAN = Name('R')
+LOWERCASE_ROMAN = Name('r')
+UPPERCASE_LETTERS = Name('A')
+LOWERCASE_LETTERS = Name('a')
+
+
+class PageLabel(Dictionary):
+    type = 'PageLabel'
+
+    def __init__(self, numbering_style=None, label_prefix=None, start=None):
+        super().__init__(indirect=False)
+        if numbering_style:
+            self['S'] = numbering_style
+        if label_prefix is not None:
+            self['P'] = String(label_prefix)
+        if start is not None:
+            self['St'] = Integer(start)
+
+
 class Outlines(Dictionary):
     type = 'Outlines'
 

src/rinoh/document.py

@@ -84,7 +84,8 @@ def __init__(self, document_part, paper, orientation=PORTRAIT):
         document = self.document_part.document
         backend_document = document.backend_document
         self.backend_page = document.backend.Page(backend_document,
-                                                  width, height)
+                                                  width, height, self.number,
+                                                  self.number_format)
         self.section = None     # will point to the last section on this page
         self.overflowed_chains = []
         self._current_section = {}

src/rinoh/style.py

@@ -653,9 +653,15 @@ def find_matches(self, styled, container):
     def find_style(self, styled, container):
         matches = sorted(self.find_matches(styled, container),
                          key=attrgetter('specificity'), reverse=True)
-        if len(set(match.specificity for match in matches)) < len(matches):
-            styled.warn('Multiple selectors match with the same specificity. '
-                        'See the style log for details.', container)
+        if len(matches) > 1:
+            last_match = matches[0]
+            for match in matches[1:]:
+                if (match.specificity == last_match.specificity
+                        and match.style_name != last_match.style_name):
+                    styled.warn('Multiple selectors match with the same '
+                                'specificity. See the style log for details.',
+                                container)
+                last_match = match
         for match in matches:
             try:
                 return self[match.style_name]

tests_regression/helpers/pdf_linkchecker.py

@@ -3,86 +3,38 @@
 # https://gist.github.com/tiarno/dea01f70a54cac52f6a6
 
 import sys
-import urllib
 
-from PyPDF2 import PdfFileReader
+from rinoh.backend.pdf import PDFReader
 
-import requests
-
-
-def check_ftp(url):
-    try:
-        response = urllib.urlopen(url)
-    except IOError as e:
-        result, reason = False, e
-    else:
-        if response.read():
-            result, reason = True, 'okay'
-        else:
-            result, reason = False, 'Empty Page'
-    return result, reason
-
-
-def check_url(url, auth=None):
-    headers = {'User-Agent': 'Mozilla/5.0', 'Accept': '*/*'}
-    if url.startswith('ftp://'):
-        result, reason = check_ftp(url)
-    else:
-        try:
-            response = requests.get(url, timeout=6, auth=auth, headers=headers)
-        except (requests.ConnectionError,
-                requests.HTTPError,
-                requests.Timeout,
-                requests.exceptions.MissingSchema,
-                requests.exceptions.InvalidSchema) as e:
-            result, reason = False, e
-        else:
-            if response.text:
-                result, reason = response.status_code, response.reason
-            else:
-                result, reason = False, 'Empty Page'
-
-    return result, reason
 
 
 def check_pdf(pdf):
     links = list()
     urls = list()
     badurls = list()
 
-    for page in pdf.pages:
-        obj = page.getObject()
-        for annot in [x.getObject() for x in obj.get('/Annots', [])]:
-            if '/A' in annot:
-                dst = annot['/A'].get('/D')
-                url = annot['/A'].get('/URI')
-                if dst:
-                    links.append(dst)
-                elif url:
-                    continue
-                    urls.append(url)
-                    result, reason = check_url(url)
-                    if not result:
-                        badurls.append({'url':url, 'reason': '%r' % reason})
-            elif '/Dest' in annot:
-                links.append(annot['/Dest'])
+    for page in pdf.catalog['Pages']['Kids']:
+        obj = page.object
+        for annot in [x.object for x in obj.get('Annots', [])]:
+            if 'A' in annot and 'D' in annot['A']:
+                links.append(str(annot['A']['D']))
+            elif 'Dest' in annot:
+                links.append(str(annot['Dest']))
 
-
-    anchors = pdf.namedDestinations.keys()
+    anchors = [str(key) for key in pdf.catalog['Names']['Dests']['Names'][::2]]
     superfluous_anchors = [x for x in anchors if x not in links]
     badlinks = [x for x in links if x not in anchors]
     return anchors, links, superfluous_anchors, badlinks, urls, badurls
 
 
 def check_pdf_links(filename):
-    pdf = PdfFileReader(filename)
+    pdf = PDFReader(filename)
     return check_pdf(pdf)
 
 
 if __name__ == '__main__':
     fname = sys.argv[1]
     print('Checking %s' % fname)
-    pdf = PdfFileReader(fname)
     (anchors, links, superfluous_anchors,
      badlinks, urls, badurls) = check_pdf_links(fname)
     print('urls: ', ', '.join(urls))

tox.ini

@@ -35,8 +35,6 @@ commands =
 deps =
     {[testenv]deps}
     pytest-assume
-    requests
-    PyPDF2
 commands =
     ./run_tests.sh {posargs} tests_regression