abiword import fails with UnicodeDecodeError

[Michael Hudson-Doyle]

The abiword import fails like this:

Traceback (most recent call last):
  File "/home/importd/dists/launchpad/lib/importd/CommandLineRunner.py", line 67, in ?
    main()
  File "/home/importd/dists/launchpad/lib/importd/CommandLineRunner.py", line 64, in main
    doer.runMethod(method, dirname)
  File "/home/importd/dists/launchpad/lib/importd/CommandLineRunner.py", line 57, in runMethod
    getattr(self._job, method)(logger=self.logger, dir=dirname)
  File "/home/importd/dists/launchpad/lib/importd/Job.py", line 188, in runJob
    strategy(self, dir, logger)
  File "/home/importd/dists/launchpad/lib/importd/JobStrategy.py", line 108, in Import
    self._importIncrementally(target_path)
  File "/home/importd/dists/launchpad/lib/importd/JobStrategy.py", line 153, in _importIncrementally
    self.runtobaz("-SC", "%s::" % lastCommit, target_path, self.logger)
  File "/home/importd/dists/launchpad/lib/importd/JobStrategy.py", line 86, in runtobaz
    totla.totla(config, logger, config.args, self.sourceTree())
  File "/home/importd/dists/launchpad/lib/cscvs/cmds/totla.py", line 263, in totla
    apply_changeset(source_tree, logger, apply, revision, target_tree, revisions)
  File "/home/importd/dists/launchpad/lib/cscvs/cmds/totla.py", line 129, in apply_changeset
    fileChange.apply(logger, target_tree, prune)
  File "/home/importd/dists/launchpad/lib/CVS/StorageLayer.py", line 912, in apply
    self._add(logger, target_tree)
  File "/home/importd/dists/launchpad/lib/CVS/StorageLayer.py", line 838, in _add
    self._change.write_file_breaking_links(target_tree, self.filename, delta, mode)
  File "/home/importd/dists/launchpad/lib/SCM/__init__.py", line 195, in write_file_breaking_links
    ",,new.%s.%s" % (random.random(), os.path.basename(filename)))
  File "/usr/lib/python2.4/posixpath.py", line 65, in join
    path += '/' + b
UnicodeDecodeError: 'ascii' codec can't decode byte 0xa7 in position 23: ordinal not in range(128)

Some kind of confusion with unicode file names?

[David Allouche (ddaa)]

Similarly, the freetrain import fails with:

  File "/home/importd/dists/launchpad/lib/cscvs/cmds/totla.py", line 252, in totla
    apply_changeset(source_tree, logger, apply, revision, target_tree, revisions)
  File "/home/importd/dists/launchpad/lib/cscvs/cmds/totla.py", line 122, in apply_changeset
    fileChange.apply(logger, target_tree, prune)
  File "/home/importd/dists/launchpad/lib/svn_oo/AddedFile.py", line 27, in apply
    self.url(), self.revision)
  File "/home/importd/dists/launchpad/lib/svn_oo/util.py", line 80, in addFile
    writeFile(target_tree, relpath, url, revision)
  File "/home/importd/dists/launchpad/lib/svn_oo/util.py", line 91, in writeFile
    filemode(url, revision))
  File "/home/importd/dists/launchpad/lib/SCM/__init__.py", line 203, in write_file_breaking_links
    os.unlink(newname)
UnicodeEncodeError: 'ascii' codec can't encode characters in position 115-117: ordinal not in range(128)

Generally, cscvs does not seem to like non-ascii names.

[Robert Collins (lifeless)]

Looks like its running in a ascii locale rather than a utf8 locale. it gets the filename from cvs fine, but cannot use it in os calls.