Modify ↓

Opened 12 years ago

Closed 12 years ago

#9714 closed defect (fixed)

JsGantt fails with ChildTickets plugin and non-numeric parent value (#%s)

Reported by: rzhidalgo@yahoo.com Owned by: Chris Nelson
Priority: normal Component: TracJsGanttPlugin
Severity: normal Keywords:
Cc: Trac Release: 0.11

Description (last modified by Ryan J Ollos)

Trac JsGantt plugin fails to generate the chart with the following log error message: --- 

2012-01-20 17:44:43,257 Trac[formatter] ERROR: Macro TracJSGanttChart(milestone=Milestone1, res=0, format=quarter) failed:
Traceback (most recent call last):
  File "build/bdist.linux-i686/egg/trac/wiki/formatter.py", line 717, in _macro_formatter
    return macro.process(args, in_paragraph=True)
  File "build/bdist.linux-i686/egg/trac/wiki/formatter.py", line 304, in process
    text = self.processor(text)
  File "build/bdist.linux-i686/egg/trac/wiki/formatter.py", line 291, in _macro_processor
    text)
  File "build/bdist.linux-i686/egg/tracjsgantt/tracjsgantt.py", line 612, in expand_macro
    tasks = self._add_tasks(options)
  File "build/bdist.linux-i686/egg/tracjsgantt/tracjsgantt.py", line 565, in _add_tasks
    self.tickets.sort(key=itemgetter('wbs'))
KeyError: 'wbs'

--- The ChildTickets plugin requires a '#' sign before the ticket number as a parent field value, which seems to be causing the problem.

Attachments (0)

Change History (21)

comment:1 Changed 12 years ago by Ryan J Ollos

Description: modified (diff)

comment:2 Changed 12 years ago by Chris Nelson

Have you set the parent_format option? It defaults to '%s' but can be set to '#%s'. Looking now, I see that the example 

# To work with ChildTickets plugin parent_format', '%s', Format of ticket IDs in parent field (default: None).
  parent_format = %s

Is rather vague and unhelpful. I believe that you want is 

[TracPM]
parent_format=#%s

in trac.ini but I have no good way to test it.

comment:3 Changed 12 years ago by Chris Nelson

Status: new → assigned

comment:4 Changed 12 years ago by rzhidalgo@yahoo.com

Yes, the parent_format was already set to '#%s'. We've been using JsGantt and ChildTickets for a while but the later version (r11133), with 'tracjsgantt.tracpm.*', started showing the 'wbs' error.

comment:5 in reply to:  4 Changed 12 years ago by Chris Nelson

Replying to rozhi:

Yes, the parent_format was already set to '#%s'. We've been using JsGantt and ChildTickets for a while but the later version (r11133), with 'tracjsgantt.tracpm.*', started showing the 'wbs' error.

OK. Thanks. I'll look into it.

comment:6 Changed 12 years ago by Chris Nelson

(In [11182]) Trying to find WBS issue with ChildTickets. Refs #9714.

Just adding logging.

comment:7 Changed 12 years ago by Chris Nelson

Please send me the lines that this change puts in trac.log so I can try to solve your problem. Thanks.

comment:8 in reply to:  7 ; Changed 12 years ago by rzhidalgo@yahoo.com

Replying to ChrisNelson:

Please send me the lines that this change puts in trac.log so I can try to solve your problem. Thanks.

Here's the log: 

2012-01-23 12:28:04,575 Trac[tracjsgantt] DEBUG: Computing WBS
2012-01-23 12:28:04,576 Trac[tracjsgantt] DEBUG: Computing for 749
2012-01-23 12:28:04,576 Trac[tracjsgantt] DEBUG:   parent is 0
2012-01-23 12:28:04,576 Trac[tracjsgantt] DEBUG:   setting level
2012-01-23 12:28:04,576 Trac[tracjsgantt] DEBUG: Computing WBS for (749, [1], 1)
2012-01-23 12:28:04,576 Trac[tracjsgantt] DEBUG: 749 has children []
2012-01-23 12:28:04,576 Trac[tracjsgantt] DEBUG: Computing for 742
2012-01-23 12:28:04,576 Trac[tracjsgantt] DEBUG:   parent is 0
2012-01-23 12:28:04,576 Trac[tracjsgantt] DEBUG:   setting level
2012-01-23 12:28:04,576 Trac[tracjsgantt] DEBUG: Computing WBS for (742, [2], 1)
2012-01-23 12:28:04,576 Trac[tracjsgantt] DEBUG: 742 has children []
2012-01-23 12:28:04,576 Trac[tracjsgantt] DEBUG: Computing for 847
2012-01-23 12:28:04,576 Trac[tracjsgantt] DEBUG:   parent is 0
2012-01-23 12:28:04,576 Trac[tracjsgantt] DEBUG:   setting level
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG: Computing WBS for (847, [3], 1)
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG: 847 has children []
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG: Computing for 949
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG:   parent is 0
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG:   setting level
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG: Computing WBS for (949, [4], 1)
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG: 949 has children [382]
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG: Computing WBS for (382, [4, 1], 2)
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG: 382 has children []
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG: Computing for 948
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG:   parent is 949
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG: Computing for 892
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG:   parent is 0
2012-01-23 12:28:04,577 Trac[tracjsgantt] DEBUG:   setting level
2012-01-23 12:28:04,578 Trac[tracjsgantt] DEBUG: Computing WBS for (892, [5], 1)
2012-01-23 12:28:04,578 Trac[tracjsgantt] DEBUG: 892 has children []
2012-01-23 12:28:04,578 Trac[tracjsgantt] DEBUG: Computing for 382
2012-01-23 12:28:04,578 Trac[tracjsgantt] DEBUG:   parent is 949
2012-01-23 12:28:04,578 Trac[tracjsgantt] DEBUG: Computing for -1
2012-01-23 12:28:04,578 Trac[tracjsgantt] DEBUG:   parent is 0
2012-01-23 12:28:04,578 Trac[tracjsgantt] DEBUG:   setting level
2012-01-23 12:28:04,578 Trac[tracjsgantt] DEBUG: Computing WBS for (-1, [6], 1)
2012-01-23 12:28:04,578 Trac[tracjsgantt] DEBUG: -1 has children None
2012-01-23 12:28:04,580 Trac[formatter] ERROR: Macro TracJSGanttChart(milestone=ATMS 3.0.0, format=quarter, lwidth=700) failed: 
Traceback (most recent call last):
  File "build/bdist.linux-i686/egg/trac/wiki/formatter.py", line 717, in _macro_formatter
    return macro.process(args, in_paragraph=True)
  File "build/bdist.linux-i686/egg/trac/wiki/formatter.py", line 304, in process
    text = self.processor(text)
  File "build/bdist.linux-i686/egg/trac/wiki/formatter.py", line 291, in _macro_processor
    text)
  File "build/bdist.linux-i686/egg/tracjsgantt/tracjsgantt.py", line 619, in expand_macro
    tasks = self._add_tasks(options)
  File "build/bdist.linux-i686/egg/tracjsgantt/tracjsgantt.py", line 572, in _add_tasks
    self.tickets.sort(key=itemgetter('wbs'))
KeyError: 'wbs'

The tickets are 749, 742, 847, 892, and 949 (children(382,948)).
The parent field has #949.
The compute_wbs is okay, but the tickets.sort is failing.
Thanks for your help.

comment:9 in reply to:  8 Changed 12 years ago by Chris Nelson

Replying to rozhi:

Replying to ChrisNelson:

Please send me the lines that this change puts in trac.log so I can try to solve your problem. Thanks.

Here's the log: ... The compute_wbs is okay, but the tickets.sort is failing.

Yes, but the sort is failing because the wbs field is missing and it should only be missing if computing the wbs failed or was skipped.

comment:10 Changed 12 years ago by Chris Nelson

The pseudoticket for the milestone (ticket ID -1) had children None.

children should only be None if no parent field is configured. But the parent field is configured because some tickets have no children ([]) and 949 has child 382.

I don't see any way for [] and None to both be used with the same configuration.

Can you post your [TracPM] section of trac.ini?

comment:11 in reply to:  10 ; Changed 12 years ago by Chris Nelson

Replying to ChrisNelson:

The pseudoticket for the milestone (ticket ID -1) had children None.

children should only be None if no parent field is configured. But the parent field is configured because some tickets have no children ([]) and 949 has child 382.

I don't see any way for [] and None to both be used with the same configuration.

Can you post your [TracPM] section of trac.ini?

Any news?

comment:12 in reply to:  11 ; Changed 12 years ago by rzhidalgo@yahoo.com

Replying to ChrisNelson:

Replying to ChrisNelson:

The pseudoticket for the milestone (ticket ID -1) had children None.

children should only be None if no parent field is configured. But the parent field is configured because some tickets have no children ([]) and 949 has child 382.

I don't see any way for [] and None to both be used with the same configuration.

Can you post your [TracPM] section of trac.ini?

Any news?

I'm trying to do some debugging also. I manually changed the Childticket parent values (removed #), and it sucessfully generate the chart. Both logs (good and failing) have [] and None. The difference is the failing code did not find all the children. Ticket 949 has children 382 and 948.
Fail (parent value with #): 

2012-01-26 15:18:55,859 Trac[tracjsgantt] DEBUG: Computing WBS for (949, [4], 1)
2012-01-26 15:18:55,859 Trac[tracjsgantt] DEBUG: 949 has children [382]

Good (removed # from parent values): 

2012-01-26 15:15:33,677 Trac[tracjsgantt] DEBUG: Computing WBS for (949, [4], 1)
2012-01-26 15:15:33,677 Trac[tracjsgantt] DEBUG: 949 has children [382, 948]
2012-01-26 15:15:33,677 Trac[tracjsgantt] DEBUG: Computing WBS for (948, [4, 1], 2)

I couldn't find the code where you strip the '#'. It used to be in tracjsgant.py. Thanks.

comment:13 in reply to:  12 Changed 12 years ago by Chris Nelson

Replying to rozhi:

... I couldn't find the code where you strip the '#'. It used to be in tracjsgant.py. Thanks.

I refactored the shim-stuff out of tracjsgantt.py to tracpm.py a few revs ago. It has: 

    def postQuery(self, options, tickets):
        # Handle custom fields.

        # Clean up custom fields which might be null ('--') vs. blank ('')
...

        # Normalize parent field values.  All parent values must be
        # done before building child lists, below.
        if self.isCfg('parent'):
            for t in tickets:
                # ChildTicketsPlugin puts '#' at the start of the
                # parent field.  Strip it for simplicity.
                fieldName = self.fields[self.sources['parent']]
                parent = t[fieldName]
                if len(parent) > 0 and parent[0] == '#':
                    t[fieldname] = parent[1:]

comment:14 Changed 12 years ago by didley@…

I'm using childticketplugin also. Is the parent field empty JSGantt works. Is parent field filled with a number for example #XXXX than it fails with error below. 

Traceback (most recent call last):
  File "C:\Programme\Python25\lib\site-packages\trac-0.12.2-py2.5.egg\trac\wiki\formatter.py", line 717, in _macro_formatter
    return macro.process(args, in_paragraph=True)
  File "C:\Programme\Python25\lib\site-packages\trac-0.12.2-py2.5.egg\trac\wiki\formatter.py", line 304, in process
    text = self.processor(text)
  File "C:\Programme\Python25\lib\site-packages\trac-0.12.2-py2.5.egg\trac\wiki\formatter.py", line 291, in _macro_processor
    text)
  File "build\bdist.win32\egg\tracjsgantt\tracjsgantt.py", line 621, in expand_macro
    tasks = self._add_tasks(options)
  File "build\bdist.win32\egg\tracjsgantt\tracjsgantt.py", line 559, in _add_tasks
    self.pm.postQuery(options, self.tickets)
  File "build\bdist.win32\egg\tracjsgantt\tracpm.py", line 643, in postQuery
    t[fieldname] = parent[1:]
NameError: global name 'fieldname' is not defined

I checked this part of coding and can't find any error. It's the part of coding in comment:#12. Do you have any idea?

didley

comment:15 Changed 12 years ago by didley@…

Sorry
I meant comment:13

comment:16 in reply to:  15 ; Changed 12 years ago by anonymous

Replying to didley@gmx.de:

Sorry
I meant comment:13

I'm getting a different error. I haven't encountered your problem.

comment:17 in reply to:  16 Changed 12 years ago by didley@…

Replying to anonymous:

Replying to didley@gmx.de:

Sorry
I meant comment:13

I'm getting a different error. I haven't encountered your problem.

I opened a new ticket. #9838

didley

comment:18 Changed 12 years ago by rzhidalgo@yahoo.com

Solved the problem. The problem occurs when the ChildTickets plugin is used with JsGantt, and a parent ticket has more than 1 children. The failure, with error message (KeyError: 'wbs'), happens because a wbs was not created for a ticket (a child).
Here is the patch: 

--- a/0.11/tracjsgantt/tracpm.py	Tue Feb 21 18:26:29 2012 -0800
+++ b/0.11/tracjsgantt/tracpm.py	Tue Feb 21 18:30:38 2012 -0800
@@ -499,6 +499,15 @@
     #
     # Milestones for the tickets are added pseudo-tickets.
     def postQuery(self, options, tickets):
+
+        for t in tickets:
+            if self.isCfg('parent'):
+                # ChildTicketsPlugin puts '#' at the start of the
+                # parent field.  Strip it for simplicity.
+                parent = t[self.fields['parent']]
+                if len(parent) > 0 and parent[0] == '#':
+                    t[self.fields['parent']] = parent[1:]
+
         for t in tickets:
             # Clean up custom fields which might be null ('--') vs. blank ('')
             nullable = [ 'pred', 'succ', 
@@ -516,11 +525,6 @@
 
             # Clean up parent field, build list of children
             if self.isCfg('parent'):
-                # ChildTicketsPlugin puts '#' at the start of the
-                # parent field.  Strip it for simplicity.
-                parent = t[self.fields['parent']]
-                if len(parent) > 0 and parent[0] == '#':
-                    t[self.fields['parent']] = parent[1:]
 
                 # An empty parent default so 0 (no such ticket)
                 if t[self.fields['parent']] == '':

I moved the filtering/removal of '#', from the parent, in the beginning of postQuery, for all the tickets. The original code had it in an inner loop, but it was failing to find all the children for a parent, which later result in wbs not being created.

comment:19 Changed 12 years ago by Chris Nelson

(In [11411]) Revert "Trying to find WBS issue with ChildTickets. Refs #9714."

This reverts commit fa61684c087bdc09be9bbd9fe04e6bb008cca190.

comment:20 Changed 12 years ago by Chris Nelson

While not linked to this ticket, there were a number of changes recently that I think resulted in an equivalent fix (that is, it takes the # clean up out of an inner loop). Please let me know if the current source works for you.

comment:21 Changed 12 years ago by Chris Nelson

Resolution: → fixed
Status: assigned → closed

Haven't heard back from the OP. Confirmed that the handling of '#' is earlier in PostQuery.

Modify Ticket

Change Properties
Set your email in Preferences
Action
as closed The owner will remain Chris Nelson.
The resolution will be deleted. Next status will be 'reopened'.

Add Comment

E-mail address and name can be saved in the Preferences.

Attachments ↑ Description ↑
 
Note: See TracTickets for help on using tickets.