[postgis-devel] Issue 35 in postgis: Comments on Postgis Functions
Dane Springmeyer
blake at hailmail.net
Mon Jul 7 14:57:19 PDT 2008
Hi Regina,
On Jul 7, 2008, at 12:29 PM, Obe, Regina wrote:
> Dane,
>
> Thanks. That would help. I've started to fix these as I go along
> adding examples.
>
> I don't know if we should get rid of issue 35 though. I think the
> intent is still the same its just the actual implementation may be
> different and I was looking at your python and your comments as
> guidelines of what is wrong with the documentation.
>
Okay, sounds good.
> Is there an issue with having more than one <term> in the
> varlistentry.
No, not at all. I was commenting on multiple <term>'s used in
situations where the same <listitem> documentation info logically
applies to different functions.
For example, with the ST_Rotate* functions:
<term>ST_RotateZ(geometry, float8)</term>
<term>ST_RotateX(geometry, float8)</term>
<term>ST_RotateY(geometry, float8)</term>
Parsing this required slightly different logic, so I was concerned
that this might be an anomaly and the functions should be given their
own full <varlistentry>. But, this seems reasonable and I'm sure any
parsing logic could accommodate this.
The main place that might need attention in terms of multiple <term>'s
are these entries:
<term>ST_xmin(box3d) ymin(box3d) zmin(box3d)</term>
<term>ST_xmax(box3d) ymax(box3d) zmax(box3d)</term>
I noticed in your updates to trunk you've added the ST_prefix. Should
they also be treated like the ST_Rotate functions?
Ultimately, your use below seems perfectly appropriate, because the
same documentation (and eventually function comments) will necessary
need to be applied to those multiple function records in the database.
> The html xsl parse just seems to put in a , when forming the html
> so I assume its okay to do that.
>
> E.g. I changed Add Point entry to:
>
> <varlistentry>
> <term>ST_AddPoint(linestring geometry, point geometry)</term>
> <term>ST_AddPoint(linestring geometry, point geometry, position
> integer)</term>
>
This looks excellent. Having the geometry type listed beside the data
type will be very helpful for beginners. When I comes to parsing the
docbook to add comments do you imagine we'll just strip the added text
(in this case 'linestring') based on its position? I noticed the entry
for ST_BdPolyFromText was one of the only functions within the docbook
(before you commenced work) that listed more explicit types, but it
uses the opposite order:
<term>ST_BdPolyFromText(text WKT, integer SRID)</term>
> <listitem>
> <para>Adds a point to a LineString before point
> <position>
> (0-based index). Third parameter can be omitted or set
> to -1 for
> appending.</para>
> <programlisting>
> --guarantee all linestrings in a table are closed
> --by adding the start point of each linestring to the end of the
> line string
> --only for those that are not closed
> UPDATE sometable
> SET the_geom = ST_AddPoint(the_geom, ST_StartPoint(the_geom))
> FROM sometable
> WHERE ST_IsClosed(the_geom) = false;
> </programlisting>
Having examples like this is going to be extremely useful. I wonder if
we could develop a simple set of 'official' sample data that they
could be run upon? Obviously the documentation could instruct the
user to create some simple sample data themselves and name the table
'sometable' so a quick copy and paste could give them example results.
But I wonder whether actually using table names, field names, etc. in
the documentation that referred to sample data would work the time
investment? What do you think?
Cheers,
Dane
> From: postgis-devel-bounces at postgis.refractions.net [mailto:postgis-devel-bounces at postgis.refractions.net
> ] On Behalf Of Dane Springmeyer
> Sent: Monday, July 07, 2008 1:59 PM
> To: PostGIS Development Discussion
> Subject: Re: [postgis-devel] Issue 35 in postgis: Comments on
> Postgis Functions
>
> Regina and Paul,
>
> I agree that a python script dependency for this step isn't ideal,
> and I'm in no way committed to the idea. It was simply an easy way
> for me to start working on the task, and I was not anticipating its
> consideration as a dependency.
>
> I propose we close this issue 35 as invalid and cook up another one
> based on Regina's sweet plan. How does that sound?
>
> More comments below...
>
>
> On Jul 6, 2008, at 1:57 AM, Obe, Regina wrote:
>
>> Paul,
>> You must have been reading my mind when you selectively picked this
>> one and the proj one to not comment about.
>>
>> Regarding this I have been looking at Dane's submission and as Paul
>> mentioned, off list, its not good to add any more dependencies than
>> we need to. This would require adding Python and Python's pgsql
>> library to use his source.
>
> Yes, it requires Python and the Psycopg2 driver as well as an
> existing db to match function names against (due to variable
> argument syntax).
>
>>
>> My main gripe with it is that it requires a postgis_template db to
>> be loaded to build the comments. It seems the main reason for that
>> is to look up the arg signature in the procname which would seem to
>> be unnecessary if our argument list in the documents could be
>> mapped to the postgresql function signature.
>
> Exactly. It could easily be modified to match against an xml file
> (or internal python data type) that stores the exact function names
> and arguments, but having the docs more closely match the actual
> function definitions would be ideal.
>
>>
>> In many cases they do or almost do. But we have cases like
>>
>> ST_MakePolygon(linestring, [linestring[]])
>>
>> which if we changed to
>>
>> ST_MakePolygon(linestring geometry, linestrings geometry[])
>> ST_MakePolygon(linestring)
>>
>> would work.
>>
>> the only issue I can think of are these functions with multi-
>> signatures
>> ST_AddPoint(linestring, point, [<position>])
>>
>> I personally would prefer see it listed twice as
>> ST_AddPoint(linestring geometry, point geometry)
>>
>> ST_AddPoint(linestring geometry, point geometry, position integer)
>>
>> since I'm not convinced non-programmers don't find that
>> nomenclature confusing.
>>
>> then I'm thinking we can generate the postgis_comments.sql.in from
>> a simple
>>
>> postgis_comments.xsl file that parses the postgis.xml thus not
>> needing another dependency.
>>
>> Although I could very well be trivializing the simplicity of this.
>>
>> Does anyone have issues with me changing the doc accordingly.
>>
>
> This sounds excellent. I'd be happy to help with this renaming if it
> is needed.
>
>
> Cheers,
>
> Dane
>
>
>
>
>> Thanks,
>> Regina
>>
>>
>>
>> From: postgis-devel-bounces at postgis.refractions.net on behalf of codesite-noreply at google.com
>> Sent: Thu 7/3/2008 5:47 PM
>> To: postgis-devel at postgis.refractions.net
>> Subject: [postgis-devel] Issue 35 in postgis: Comments on Postgis
>> Functions
>>
>> Issue 35: Comments on Postgis Functions
>> http://code.google.com/p/postgis/issues/detail?id=35
>>
>> Comment #2 by pwramsey3:
>> (No comment was entered for this change.)
>>
>>
>> Issue attribute updates:
>> Labels: -Type-Defect Type-Enhancement
>>
>> --
>> You received this message because you are listed in the owner
>> or CC fields of this issue, or because you starred this issue.
>> You may adjust your issue notification preferences at:
>> http://code.google.com/hosting/settings
>> _______________________________________________
>> postgis-devel mailing list
>> postgis-devel at postgis.refractions.net
>> http://postgis.refractions.net/mailman/listinfo/postgis-devel
>>
>>
>>
>> The substance of this message, including any attachments, may be
>> confidential, legally privileged and/or exempt from disclosure
>> pursuant to Massachusetts law. It is intended solely for the
>> addressee. If you received this in error, please contact the sender
>> and delete the material from any computer.
>>
>>
>>
>> Help make the earth a greener place. If at all possible resist
>> printing this email and join us in saving paper.
>>
>>
>> _______________________________________________
>> postgis-devel mailing list
>> postgis-devel at postgis.refractions.net
>> http://postgis.refractions.net/mailman/listinfo/postgis-devel
>
>
>
> The substance of this message, including any attachments, may be
> confidential, legally privileged and/or exempt from disclosure
> pursuant to Massachusetts law. It is intended solely for the
> addressee. If you received this in error, please contact the sender
> and delete the material from any computer.
>
>
>
> Help make the earth a greener place. If at all possible resist
> printing this email and join us in saving paper.
>
>
> _______________________________________________
> postgis-devel mailing list
> postgis-devel at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-devel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/postgis-devel/attachments/20080707/bdece57f/attachment.html>
More information about the postgis-devel
mailing list