Closed (duplicate)
Project:
Drupal core
Version:
7.x-dev
Component:
documentation
Priority:
Normal
Category:
Bug report
Assigned:
Unassigned
Reporter:
Created:
1 Jan 2011 at 17:51 UTC
Updated:
14 Oct 2011 at 21:38 UTC
Jump to comment: Most recent file
Comments
Comment #1
tstoecklerSetting to "needs review".
Comment #2
jhodgdonThis looks pretty good, thanks!
I can suggest a few more improvements:
a)
Should be Initializes the PHP...
b)
This could maybe use more elaboration - as a one-line description, "loads the configuration" leaves me asking: what configuration? Maybe "Loads the site configuration from settings.php" or whatever it does. Also, probably "correctly" is not necessary at the end. Remember, these one-line descriptions are used for many things on api.drupal.org and are not always seen in the context of the file they're in, etc.
c)
Not very good wrapping - first line should be longer.
d)
No comma before "that"
e)
be -> been
f) You might just leave the t() documentation alone, as it is being completely reworked in another issue:
#997884: t() documentation overhaul
g) Regarding those boostrap phase functions, one thing we could do is make them, e.g.:
That would be more in compliance with our standards than the current state of the doc headers.
Comment #3
tstoecklerFixed #2. Are the function descriptions for the bootstrap hooks OK? I also changed drupal_static()'s one-line description to "Stores static variables centrally." (was: "Central static variable storage").
Comment #4
tstoecklerAnd now with something to actually review...
Comment #5
jhodgdonLooking better! I see a couple more things that could be improved (these are not things you necessarily introduced, but things that should eventually be fixed):
a)
"to get cached" should be "from being cached".
b)
Since this is a true/false, and aside from the variable name $append it's not all that clear whether "append" or "replace" would correspond to true/false, I would prefer the wording: "TRUE to append the value to an existing header, or FALSE to replace the header." I think it's clearer.
c)
There should not be a comma before "that"
d)
$obj should be renamed to $object, to comply with our coding standards. Since this is a local variable, it's considered a doc fix not an API change to change the param name for clarity.
e) Please just take out any changes to t() documentation and leave them out of this patch. They're being discussed elsewhere (see above).
f)
first line: needs a comma before which
last two lines: "a text" should either be "text" or "a string", and passed-in should be hyphenated.
g)
The first or needs a comma before it, I think, for clarity.
h)
which -> that
i)
What's a hmac? If it's an acronym, should it be HMAC? I see it is HMAC down further.
j)
no comma before or
Comment #6
tstoecklerAnother try.
Comment #7
jhodgdonItem (c) from comment #2 above is still there. So is item (a), and item (b) is not fixed either.... actually it looks like all the items in #2 are back.
t() doc changes are still in there too.
And I just noticed one more verb tense:
Define -> Defines
Comment #8
tstoecklerPicked this up after having forgotten it for a long time.
I double-checked that this fixed everything in #2, #5 and #7 (I don't know what had happened with the last patch). I also fixed a lot of inconsistent wrapping in the file. A couple of times things would wrap after 80 chars but most of the time too early. That includes one wrapping change in the t()-documentation, but since the t()-docs-patch was already committed that shouldn't cause any problem.
Crossing my fingers that this will go through... :)
Comment #9
tstoecklerneeds review
Comment #10
jhodgdonThis is pretty good. One error:
The word "at" should not be removed here, I think?
Gracious, the doc in that function is a mess of grammar. I guess I'll file a separate issue on that. For now probably just replace the "at" where it was. Here's the separate issue:
#1071846: conf_path() doc needs cleanup
I don't have time to review the rest of this patch at the moment (got about 30% done).
Comment #11
tstoecklerRerolled. I kept the incorrect wrapping in conf_path() for now. If #1071846: conf_path() doc needs cleanup gets committed first I'll reroll (or that one, if this gets in first).
Comment #12
tstoecklerRerolled. I kept the incorrect wrapping in conf_path() for now. If #1071846: conf_path() doc needs cleanup gets committed first I'll reroll (or that one, if this gets in first).
Comment #14
jhodgdonWe really should not do such big patches -- there's always some little problem and it derails the whole patch...
a)
Missing . at the end of the line. The same problem is in the next few functions as well, and here:
b) Inconsistent use of SimpleTest vs. simpletest:
It seems to be "simpletest" everywhere else in this file, so let's stick with that.
c) Extra hunks at the end of the patch file: aggregator.info and block.info
Comment #15
tstoecklerThat's why I would advocate to commit some form of this, and then handle the rest in follow-up patches. That's kind of why people invented version control...
I'll reroll in the next couple of days.
c) oops.., sorry.
Comment #16
jhodgdonWell how about just making a problem-free patch for part of the file, and then a follow-up patch with another part? I don't really like to mark a patch as RTBC if it is fixing one error only to replace it with another error.
Comment #17
jhodgdonWe're now doing this kind of cleanup on #1310084: [meta] API documentation cleanup sprint in larger batches.