#7177 Trac ticket error: astimezone() cannot be applied to a naive datetime

v1.2.0
closed
nobody
trac (14) import (116)
General
Tim Van Steenburgh
2015-08-20
2014-02-14
Dave Brondsema
No

Errored on a CSV file like this from a Trac 1.0.1 instance:

__group__,ticket,summary,component,status,type,priority,owner,modified,_time,_reporter
prep,48,Create Stuff,Application,new,build,major,nobody,2014-02-09T05:16:55-05:00,16:14:18-05:00,John Doe

Trace:

Traceback (most recent call last):
  File "/var/local/allura/ForgeImporters/forgeimporters/base.py", line 131, in import_tool
    mount_point=mount_point, mount_label=mount_label, **kw)
  File "/var/local/allura/ForgeImporters/forgeimporters/trac/tickets.py", line 119, in import_tool
    json.dumps(export(trac_url), cls=DateJSONEncoder),
  File "/var/local/allura/Allura/allura/scripts/trac_export.py", line 304, in export
    verbose=verbose, do_attachments=do_attachments)
  File "/var/local/allura/Allura/allura/scripts/trac_export.py", line 99, in __init__
    self.ticket_queue = self.next_ticket_ids()
  File "/var/local/allura/Allura/allura/scripts/trac_export.py", line 263, in next_ticket_ids
    r[1]), 'date_updated': self.trac2z_date(r[2])}
  File "/var/local/allura/Allura/allura/scripts/trac_export.py", line 127, in trac2z_date
    d = d.astimezone(pytz.UTC)
ValueError: astimezone() cannot be applied to a naive datetime

Discussion

  • Anonymous - 2014-02-16

    Originally by: yyachim

    In my case, Trac 1.0.1 produced timestamps that excluded the timezone designator (i.e. +/-hh:mm) in the CSV file. Skipping the call to astimezone() in seemed to work:

        if d.tzinfo is not None:
        d = d.astimezone(pytz.UTC)

    Unrelated to this ticket, but another issue with Trac 1.0.1, is the byte-order mark in the beginning of the CSV file. For me, this caused extra bytes to be included as part of the 'id' key of a dictionary in remap_fields(). I was able to export my data after updating that function to this:

        def remap_fields(self, dict):
        "Remap fields to adhere to standard taxonomy."
        out = {}
        for k, v in dict.iteritems():
            key = self.match_pattern(r'\W*(\w+)\W*', k)
            out[self.FIELD_MAP.get(key, key)] = v

        out['id'] = int(out['id'])
        if 'private' in out:
            out['private'] = bool(int(out['private']))
        return out
     
    allura_ticket_7177_yyachim-1.patch
    software_cracking.pdf

  • Dave Brondsema

    Dave Brondsema - 2014-02-19
    • status: open --> code-review
     

  • Tim Van Steenburgh - 2014-02-20
    • QA: Tim Van Steenburgh
    • Milestone: forge-backlog --> forge-feb-21
     

  • Tim Van Steenburgh - 2014-02-20
    • status: code-review --> closed
     

  • Tim Van Steenburgh - 2014-02-20

    Nice, thanks for the patch!

     

  • Dave Brondsema

    Dave Brondsema - 2014-02-24

    FYI - this is now live on sourceforge.net

     

  • Anonymous - 2014-02-25

    Originally by: yyachim

    Thanks for the quick turnaround.

     
    software_cracking.pdf

  • Dave Brondsema

    Dave Brondsema - 2015-01-05
    • Milestone: unreleased --> asf_release_1.2.0
     

Log in to post a comment.