D5767: subrepo: clean up lingering bytes/str issues in svn support

[durin42 (Augie Fackler)]

durin42 created this revision.
Herald added a reviewer: pulkit.
Herald added a subscriber: mercurial-devel.
Herald added a reviewer: hg-reviewers.

REVISION SUMMARY
  I've decided to use fsencode instead of utf8 for turning Subversion
  paths into native paths. I don't really know which one is more
  correct, but I _suspect_ that this means that Python3 will avoid some
  bugs that are possible on Python2 when it comes to Subversion
  functionality.
  
  Subversion subrepo tests now pass in Python 3.

REPOSITORY
  rHG Mercurial

REVISION DETAIL
  https://phab.mercurial-scm.org/D5767

AFFECTED FILES
  contrib/python3-whitelist
  mercurial/subrepo.py

CHANGE DETAILS

diff --git a/mercurial/subrepo.py b/mercurial/subrepo.py
--- a/mercurial/subrepo.py
+++ b/mercurial/subrepo.py
@@ -961,6 +961,7 @@
                              universal_newlines=True,
                              env=procutil.tonativeenv(env), **extrakw)
         stdout, stderr = p.communicate()
+        stdout, stderr = pycompat.fsencode(stdout), pycompat.fsencode(stderr)
         stderr = stderr.strip()
         if not failok:
             if p.returncode:
@@ -987,13 +988,14 @@
         # both. We used to store the working directory one.
         output, err = self._svncommand(['info', '--xml'])
         doc = xml.dom.minidom.parseString(output)
-        entries = doc.getElementsByTagName('entry')
+        entries = doc.getElementsByTagName(r'entry')
         lastrev, rev = '0', '0'
         if entries:
-            rev = str(entries[0].getAttribute('revision')) or '0'
-            commits = entries[0].getElementsByTagName('commit')
+            rev = pycompat.bytestr(entries[0].getAttribute(r'revision')) or '0'
+            commits = entries[0].getElementsByTagName(r'commit')
             if commits:
-                lastrev = str(commits[0].getAttribute('revision')) or '0'
+                lastrev = pycompat.bytestr(
+                    commits[0].getAttribute(r'revision')) or '0'
         return (lastrev, rev)
 
     def _wcrev(self):
@@ -1008,19 +1010,19 @@
         output, err = self._svncommand(['status', '--xml'])
         externals, changes, missing = [], [], []
         doc = xml.dom.minidom.parseString(output)
-        for e in doc.getElementsByTagName('entry'):
-            s = e.getElementsByTagName('wc-status')
+        for e in doc.getElementsByTagName(r'entry'):
+            s = e.getElementsByTagName(r'wc-status')
             if not s:
                 continue
-            item = s[0].getAttribute('item')
-            props = s[0].getAttribute('props')
-            path = e.getAttribute('path')
-            if item == 'external':
+            item = s[0].getAttribute(r'item')
+            props = s[0].getAttribute(r'props')
+            path = pycompat.fsencode(e.getAttribute(r'path'))
+            if item == r'external':
                 externals.append(path)
-            elif item == 'missing':
+            elif item == r'missing':
                 missing.append(path)
-            if (item not in ('', 'normal', 'unversioned', 'external')
-                or props not in ('', 'none', 'normal')):
+            if (item not in (r'', r'normal', r'unversioned', r'external')
+                or props not in (r'', r'none', r'normal')):
                 changes.append(path)
         for path in changes:
             for ext in externals:
@@ -1141,14 +1143,14 @@
         output = self._svncommand(['list', '--recursive', '--xml'])[0]
         doc = xml.dom.minidom.parseString(output)
         paths = []
-        for e in doc.getElementsByTagName('entry'):
-            kind = pycompat.bytestr(e.getAttribute('kind'))
+        for e in doc.getElementsByTagName(r'entry'):
+            kind = pycompat.bytestr(e.getAttribute(r'kind'))
             if kind != 'file':
                 continue
-            name = ''.join(c.data for c
-                           in e.getElementsByTagName('name')[0].childNodes
-                           if c.nodeType == c.TEXT_NODE)
-            paths.append(name.encode('utf-8'))
+            name = r''.join(c.data for c
+                            in e.getElementsByTagName(r'name')[0].childNodes
+                            if c.nodeType == c.TEXT_NODE)
+            paths.append(pycompat.fsencode(name))
         return paths
 
     def filedata(self, name, decode):
diff --git a/contrib/python3-whitelist b/contrib/python3-whitelist
--- a/contrib/python3-whitelist
+++ b/contrib/python3-whitelist
@@ -430,6 +430,7 @@
 test-mq-qrename.t
 test-mq-qsave.t
 test-mq-safety.t
+test-mq-subrepo-svn.t
 test-mq-subrepo.t
 test-mq-symlinks.t
 test-mq.t
@@ -682,6 +683,7 @@
 test-subrepo-paths.t
 test-subrepo-recursion.t
 test-subrepo-relative-path.t
+test-subrepo-svn.t
 test-subrepo.t
 test-symlink-os-yes-fs-no.py
 test-symlink-placeholder.t



To: durin42, pulkit, #hg-reviewers
Cc: mercurial-devel