Edit Book: Check Book: Add a check for "invalid" ids

Some sad-sack ebook services barf on "invalid" ids. See #1467765 (Private bug)
They should just accept all valid HTML 5 ids instead of relying
on the terrible epubcheck.

src/calibre/ebooks/oeb/polish/check/parsing.py

@@ -328,6 +328,34 @@ class DuplicateId(BaseError):
         container.dirty(self.name)
         return True
 
+class InvalidId(BaseError):
+
+    INDIVIDUAL_FIX = _(
+        'Replace this id with a randomly generated valid id')
+
+    def __init__(self, name, line, eid):
+        BaseError.__init__(self, _('Invalid id: %s') % eid, name, line)
+        self.HELP = _(
+            'The id {0} is not a valid id. IDs must start with a letter ([A-Za-z]) and may be'
+            ' followed by any number of letters, digits ([0-9]), hyphens ("-"), underscores ("_")'
+            ', colons (":"), and periods ("."). This is to ensure maximum compatibility'
+            ' with a wide range of devices.').format(eid)
+        self.invalid_id = eid
+
+    def __call__(self, container):
+        import uuid
+        from calibre.ebooks.oeb.polish.replace import replace_ids
+        newid = 'g' + uuid.uuid4().hex
+        changed = False
+        elems = (e for e in container.parsed(self.name).xpath('//*[@id]') if e.get('id') == self.invalid_id)
+        for e in elems:
+            e.set('id', newid)
+            changed = True
+            container.dirty(self.name)
+        if changed:
+            replace_ids(container, {self.name:{self.invalid_id:newid}})
+        return changed
+
 class BareTextInBody(BaseError):
 
     INDIVIDUAL_FIX = _('Wrap the bare text in a p tag')
@@ -416,6 +444,8 @@ def check_filenames(container):
             errors.append(EscapedName(name))
     return errors
 
+valid_id = re.compile(r'^[a-zA-Z][a-zA-Z0-9_:.-]*$')
+
 def check_ids(container):
     errors = []
     mts = set(OEB_DOCS) | {guess_type('a.opf'), guess_type('a.ncx')}
@@ -432,6 +462,8 @@ def check_ids(container):
                     dups[eid].append(elem.sourceline)
                 else:
                     seen_ids[eid] = elem.sourceline
+                if eid and valid_id.match(eid) is None:
+                    errors.append(InvalidId(name, elem.sourceline, eid))
             errors.extend(DuplicateId(name, eid, locs) for eid, locs in dups.iteritems())
     return errors
 

src/calibre/ebooks/oeb/polish/replace.py

@@ -8,7 +8,7 @@ __copyright__ = '2013, Kovid Goyal <kovid at kovidgoyal.net>'
 __docformat__ = 'restructuredtext en'
 
 import codecs, shutil, os, posixpath
-from urlparse import urlparse
+from urlparse import urlparse, urlunparse
 from collections import Counter, defaultdict
 
 from calibre import sanitize_file_name_unicode
@@ -46,6 +46,35 @@ class LinkReplacer(object):
             self.replaced = True
         return href
 
+class IdReplacer(object):
+
+    def __init__(self, base, container, id_map):
+        self.base, self.container, self.replaced = base, container, False
+        self.id_map = id_map
+
+    def __call__(self, url):
+        if url and url.startswith('#'):
+            repl = self.id_map.get(self.base, {}).get(url[1:])
+            if repl is None or repl == url[1:]:
+                return url
+            self.replaced = True
+            return '#' + repl
+        name = self.container.href_to_name(url, self.base)
+        if not name:
+            return url
+        id_map = self.id_map.get(name)
+        if id_map is None:
+            return url
+        purl = urlparse(url)
+        nfrag = id_map.get(purl.fragment)
+        if nfrag is None:
+            return url
+        purl = purl._replace(fragment=nfrag)
+        href = urlunparse(purl)
+        if href != url:
+            self.replaced = True
+        return href
+
 class LinkRebaser(object):
 
     def __init__(self, container, old_name, new_name):
@@ -88,6 +117,22 @@ def replace_links(container, link_map, frag_map=lambda name, frag:frag, replace_
         repl = LinkReplacer(name, container, link_map, frag_map)
         container.replace_links(name, repl)
 
+def replace_ids(container, id_map):
+    '''
+    Replace all links in the container that pointed to the changed ids.
+
+    :param id_map: A mapping of {name:id_map} where each id_map is a mapping of {old_id:new_id}
+    :return: True iff at least one link was changed
+
+    '''
+    changed = False
+    for name, media_type in container.mime_map.iteritems():
+        repl = IdReplacer(name, container, id_map)
+        container.replace_links(name, repl)
+        if repl.replaced:
+            changed = True
+    return changed
+
 def smarten_punctuation(container, report):
     from calibre.ebooks.conversion.preprocess import smarten_punctuation
     smartened = False