Re: [HACKERS] [PATCH] Revive line type
On Wed, Oct 9, 2013 at 10:44 AM, Peter Eisentraut pete...@gmx.net wrote: On Thu, 2013-10-03 at 17:50 +0530, Jeevan Chalke wrote: Will you please attach new patch with above block removed ? Then I will quickly check that new patch and mark as Ready For Committer. Here you go. Thanks Peter. Patch looks good to me and ready to go in now. Marking as Ready For Committer. -- Jeevan B Chalke Principal Software Engineer, Product Development EnterpriseDB Corporation
Re: [HACKERS] [PATCH] Revive line type
On Thu, 2013-10-03 at 17:50 +0530, Jeevan Chalke wrote: Will you please attach new patch with above block removed ? Then I will quickly check that new patch and mark as Ready For Committer. Here you go. From ba421c778cc3f7c32886ac038389cfbad3c0df67 Mon Sep 17 00:00:00 2001 From: Peter Eisentraut pete...@gmx.net Date: Wed, 9 Oct 2013 01:09:18 -0400 Subject: [PATCH v2] Revive line type MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Change the input/output format to {A,B,C}, to match the internal representation. Complete the implementations of line_in, line_out, line_recv, line_send. Remove comments and error messages about the line type not being implemented. Add regression tests for existing line operators and functions. Reviewed-by: rui hua 365507506...@gmail.com Reviewed-by: Álvaro Herrera alvhe...@2ndquadrant.com Reviewed-by: Jeevan Chalke jeevan.cha...@enterprisedb.com --- doc/src/sgml/datatype.sgml | 42 - doc/src/sgml/func.sgml | 6 + src/backend/utils/adt/geo_ops.c| 224 ++-- src/include/catalog/pg_type.h | 3 +- src/include/utils/geo_decls.h | 7 - src/test/regress/expected/geometry.out | 3 - src/test/regress/expected/line.out | 271 + src/test/regress/expected/sanity_check.out | 3 +- src/test/regress/output/misc.source| 3 +- src/test/regress/parallel_schedule | 2 +- src/test/regress/serial_schedule | 1 + src/test/regress/sql/geometry.sql | 4 - src/test/regress/sql/line.sql | 87 + 13 files changed, 503 insertions(+), 153 deletions(-) create mode 100644 src/test/regress/expected/line.out create mode 100644 src/test/regress/sql/line.sql diff --git a/doc/src/sgml/datatype.sgml b/doc/src/sgml/datatype.sgml index 87668ea..07f0385 100644 --- a/doc/src/sgml/datatype.sgml +++ b/doc/src/sgml/datatype.sgml @@ -3051,9 +3051,7 @@ titleGeometric Types/title para Geometric data types represent two-dimensional spatial objects. xref linkend=datatype-geo-table shows the geometric -types available in productnamePostgreSQL/productname. The -most fundamental type, the point, forms the basis for all of the -other types. +types available in productnamePostgreSQL/productname. /para table id=datatype-geo-table @@ -3063,8 +3061,8 @@ titleGeometric Types/title row entryName/entry entryStorage Size/entry -entryRepresentation/entry entryDescription/entry +entryRepresentation/entry /row /thead tbody @@ -3077,8 +3075,8 @@ titleGeometric Types/title row entrytypeline/type/entry entry32 bytes/entry -entryInfinite line (not fully implemented)/entry -entry((x1,y1),(x2,y2))/entry +entryInfinite line/entry +entry{A,B,C}/entry /row row entrytypelseg/type/entry @@ -3153,6 +3151,38 @@ titlePoints/title /sect2 sect2 +titleLines/title + +indexterm + primaryline/primary +/indexterm + +para + Lines (typeline/type) are represented by the linear equation Ax + By + + C = 0, where A and B are not both zero. Values of + type typeline/type is input and output in the following form: +synopsis +{ replaceableA/replaceable, replaceableB/replaceable, replaceableC/replaceable } +/synopsis + + Alternatively, any of the following forms can be used for input: + +synopsis +[ ( replaceablex1/replaceable , replaceabley1/replaceable ) , ( replaceablex2/replaceable , replaceabley2/replaceable ) ] +( ( replaceablex1/replaceable , replaceabley1/replaceable ) , ( replaceablex2/replaceable , replaceabley2/replaceable ) ) + ( replaceablex1/replaceable , replaceabley1/replaceable ) , ( replaceablex2/replaceable , replaceabley2/replaceable ) +replaceablex1/replaceable , replaceabley1/replaceable , replaceablex2/replaceable , replaceabley2/replaceable +/synopsis + + where + literal(replaceablex1/replaceable,replaceabley1/replaceable)/literal + and + literal(replaceablex2/replaceable,replaceabley2/replaceable)/literal + are two (different) points on the line. +/para + /sect2 + + sect2 titleLine Segments/title indexterm diff --git a/doc/src/sgml/func.sgml b/doc/src/sgml/func.sgml index 7dd1ef2..b162618 100644 --- a/doc/src/sgml/func.sgml +++ b/doc/src/sgml/func.sgml @@ -8123,6 +8123,12 @@ titleGeometric Type Conversion Functions/title entryliteralcircle(polygon '((0,0),(1,1),(2,0))')/literal/entry /row row +entryliteralfunctionline(typepoint/type, typepoint/type)/function/literal/entry +entrytypeline/type/entry +entrypoints to line/entry +entryliteralline(point '(-1,0)', point '(1,0)')/literal/entry + /row + row entry
Re: [HACKERS] [PATCH] Revive line type
On Wed, Oct 2, 2013 at 6:12 AM, Peter Eisentraut pete...@gmx.net wrote: On Wed, 2013-09-25 at 14:26 +0530, Jeevan Chalke wrote: So no issues from my side. However, do we still need this in close_pl() ? #ifdef NOT_USED if (FPeq(line-A, -1.0) FPzero(line-B)) {/* vertical */ } #endif No, that can be removed. Will you please attach new patch with above block removed ? Then I will quickly check that new patch and mark as Ready For Committer. Also close_sl, close_lb and dist_lb are NOT yet implemented. It will be good if we have those. But I don't think we should wait for those functions to be implemented. Right, those are separate projects. Agree. Thanks -- Jeevan B Chalke
Re: [HACKERS] [PATCH] Revive line type
On Wed, 2013-09-25 at 14:26 +0530, Jeevan Chalke wrote: So no issues from my side. However, do we still need this in close_pl() ? #ifdef NOT_USED if (FPeq(line-A, -1.0) FPzero(line-B)) {/* vertical */ } #endif No, that can be removed. Also close_sl, close_lb and dist_lb are NOT yet implemented. It will be good if we have those. But I don't think we should wait for those functions to be implemented. Right, those are separate projects. -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
Re: [HACKERS] [PATCH] Revive line type
Hi, I had a look over this patch and here are my review points: 1. Patch applies cleanly. 2. make, make install and make check is good. 3. I did lot of random testing and didn't find any issue. 4. Test coverage is very well. It has all scenarios and all operators are tested with line. That's really great. So no issues from my side. However, do we still need this in close_pl() ? #ifdef NOT_USED if (FPeq(line-A, -1.0) FPzero(line-B)) {/* vertical */ } #endif Also close_sl, close_lb and dist_lb are NOT yet implemented. It will be good if we have those. But I don't think we should wait for those functions to be implemented. We can go ahead with this. Please confirm above concern so that I will mark it as Ready for Committer. Thanks On Tue, Sep 17, 2013 at 7:34 AM, Alvaro Herrera alvhe...@2ndquadrant.comwrote: David Fetter escribió: Should the things you tried and others be in the regression tests? If so, should we start with whatever had been in the regression tests when the line type was dropped? Actually, the patch does include a regression test for the revived type (and it passes). I don't think more than that is needed. -- Álvaro Herrerahttp://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Training Services -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers -- Jeevan B Chalke Principal Software Engineer, Product Development EnterpriseDB Corporation
Re: [HACKERS] [PATCH] Revive line type
Peter Eisentraut escribió: Here is a new patch for the line type, with a new input/output format {A,B,C}, as discussed in this thread. I gave this a quick look. The new input/output format appears to work well. The input format using two points still works, which is nice. Regression tests pass. I tried binary input and output using COPY and it seems to round-trip without problem. With the two-point input, horizontal and vertical lines can accept NaNs without producing a NaN in the output if not necessary. I tried interpt and intersect functions and they seem to work. Haven't gotten around to trying other functions or operators. -- Álvaro Herrerahttp://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Training Services -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
Re: [HACKERS] [PATCH] Revive line type
On Mon, Sep 16, 2013 at 07:04:32PM -0300, Alvaro Herrera wrote: Peter Eisentraut escribió: Here is a new patch for the line type, with a new input/output format {A,B,C}, as discussed in this thread. I gave this a quick look. The new input/output format appears to work well. The input format using two points still works, which is nice. Regression tests pass. I tried binary input and output using COPY and it seems to round-trip without problem. With the two-point input, horizontal and vertical lines can accept NaNs without producing a NaN in the output if not necessary. I tried interpt and intersect functions and they seem to work. Haven't gotten around to trying other functions or operators. Should the things you tried and others be in the regression tests? If so, should we start with whatever had been in the regression tests when the line type was dropped? Cheers, David. -- David Fetter da...@fetter.org http://fetter.org/ Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter Skype: davidfetter XMPP: david.fet...@gmail.com iCal: webcal://www.tripit.com/feed/ical/people/david74/tripit.ics Remember to vote! Consider donating to Postgres: http://www.postgresql.org/about/donate -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
Re: [HACKERS] [PATCH] Revive line type
David Fetter escribió: Should the things you tried and others be in the regression tests? If so, should we start with whatever had been in the regression tests when the line type was dropped? Actually, the patch does include a regression test for the revived type (and it passes). I don't think more than that is needed. -- Álvaro Herrerahttp://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Training Services -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
Re: [HACKERS] [PATCH] Revive line type
Here is a new patch for the line type, with a new input/output format {A,B,C}, as discussed in this thread. From 837fcf5d9b1ee8e589ef4b19f7d6e575229ca758 Mon Sep 17 00:00:00 2001 From: Peter Eisentraut pete...@gmx.net Date: Sun, 15 Sep 2013 00:02:06 -0400 Subject: [PATCH] Revive line type Change the input/output format to {A,B,C}, to match the internal representation. Complete the implementations of line_in, line_out, line_recv, line_send. Remove comments and error messages about the line type not being implemented. Add regression tests for existing line operators and functions. Reviewed-by: rui hua 365507506...@gmail.com --- doc/src/sgml/datatype.sgml | 42 - doc/src/sgml/func.sgml | 6 + src/backend/utils/adt/geo_ops.c| 219 ++- src/include/catalog/pg_type.h | 3 +- src/include/utils/geo_decls.h | 7 - src/test/regress/expected/geometry.out | 3 - src/test/regress/expected/line.out | 271 + src/test/regress/expected/sanity_check.out | 3 +- src/test/regress/output/misc.source| 3 +- src/test/regress/parallel_schedule | 2 +- src/test/regress/serial_schedule | 1 + src/test/regress/sql/geometry.sql | 4 - src/test/regress/sql/line.sql | 87 + 13 files changed, 503 insertions(+), 148 deletions(-) create mode 100644 src/test/regress/expected/line.out create mode 100644 src/test/regress/sql/line.sql diff --git a/doc/src/sgml/datatype.sgml b/doc/src/sgml/datatype.sgml index 87668ea..07f0385 100644 --- a/doc/src/sgml/datatype.sgml +++ b/doc/src/sgml/datatype.sgml @@ -3051,9 +3051,7 @@ titleGeometric Types/title para Geometric data types represent two-dimensional spatial objects. xref linkend=datatype-geo-table shows the geometric -types available in productnamePostgreSQL/productname. The -most fundamental type, the point, forms the basis for all of the -other types. +types available in productnamePostgreSQL/productname. /para table id=datatype-geo-table @@ -3063,8 +3061,8 @@ titleGeometric Types/title row entryName/entry entryStorage Size/entry -entryRepresentation/entry entryDescription/entry +entryRepresentation/entry /row /thead tbody @@ -3077,8 +3075,8 @@ titleGeometric Types/title row entrytypeline/type/entry entry32 bytes/entry -entryInfinite line (not fully implemented)/entry -entry((x1,y1),(x2,y2))/entry +entryInfinite line/entry +entry{A,B,C}/entry /row row entrytypelseg/type/entry @@ -3153,6 +3151,38 @@ titlePoints/title /sect2 sect2 +titleLines/title + +indexterm + primaryline/primary +/indexterm + +para + Lines (typeline/type) are represented by the linear equation Ax + By + + C = 0, where A and B are not both zero. Values of + type typeline/type is input and output in the following form: +synopsis +{ replaceableA/replaceable, replaceableB/replaceable, replaceableC/replaceable } +/synopsis + + Alternatively, any of the following forms can be used for input: + +synopsis +[ ( replaceablex1/replaceable , replaceabley1/replaceable ) , ( replaceablex2/replaceable , replaceabley2/replaceable ) ] +( ( replaceablex1/replaceable , replaceabley1/replaceable ) , ( replaceablex2/replaceable , replaceabley2/replaceable ) ) + ( replaceablex1/replaceable , replaceabley1/replaceable ) , ( replaceablex2/replaceable , replaceabley2/replaceable ) +replaceablex1/replaceable , replaceabley1/replaceable , replaceablex2/replaceable , replaceabley2/replaceable +/synopsis + + where + literal(replaceablex1/replaceable,replaceabley1/replaceable)/literal + and + literal(replaceablex2/replaceable,replaceabley2/replaceable)/literal + are two (different) points on the line. +/para + /sect2 + + sect2 titleLine Segments/title indexterm diff --git a/doc/src/sgml/func.sgml b/doc/src/sgml/func.sgml index ee1c957..8f60c56 100644 --- a/doc/src/sgml/func.sgml +++ b/doc/src/sgml/func.sgml @@ -8123,6 +8123,12 @@ titleGeometric Type Conversion Functions/title entryliteralcircle(polygon '((0,0),(1,1),(2,0))')/literal/entry /row row +entryliteralfunctionline(typepoint/type, typepoint/type)/function/literal/entry +entrytypeline/type/entry +entrypoints to line/entry +entryliteralline(point '(-1,0)', point '(1,0)')/literal/entry + /row + row entry indexterm primarylseg/primary diff --git a/src/backend/utils/adt/geo_ops.c b/src/backend/utils/adt/geo_ops.c index 5d0b596..6bfe6d7 100644 --- a/src/backend/utils/adt/geo_ops.c +++ b/src/backend/utils/adt/geo_ops.c @@ -926,42 +926,82 @@ box_diagonal(PG_FUNCTION_ARGS)
Re: [HACKERS] [PATCH] Revive line type
On 6/26/13 9:34 PM, Peter Eisentraut wrote: Still wondering whether to use a A,B,C-based output format per Tom's comment. Wouldn't it also be helpful to remove The points used in the output are not necessarily the points used on input by making that not true? There are three obvious ways you might output a line: -Math class expectations of slope-intercept form: y = mx + b. Intercept forms don't handle both horizontal and vertical lines though, so that's easily out. -Pair of points. A casual observer might get lucky and observe putting a pair of points in and getting the same pair back again, then incorrectly assume that's normally the case. Seems better to never make that possible in the infinite line case. I'm reminded of how row output usually is in order gives a bad impression about ordering guarantees in SQL. -Ax+By+C=0 Outputting that third one, when it's also the internal form, handles any time of line; will show any assumptions about individual points being preserved are wrong; and avoids rounding errors too. The only downside seems to be that bounded lines are easier to show with a pair of points. I think that suggests an alternate, secondary output format would be useful though, rather than that a different default one is a good idea. -- Greg Smith 2ndQuadrant USg...@2ndquadrant.com Baltimore, MD PostgreSQL Training, Services, and 24x7 Support www.2ndQuadrant.com -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
Re: [HACKERS] [PATCH] Revive line type
On Sun, 2013-06-23 at 12:24 +0800, rui hua wrote: Regression tests are all succeed, but several problems have be found while ding some simple test. The updated document said that the points used in the output are not necessarily the points used on input. I understand that as long as they are marked on the same line. But when the input data represents a horizontal or vertical line, the output is not exactly the same line. It is another line parallel to it. postgres=# select line('1,3,2,3'); line - [(0,-3),(1,-3)] (1 row) This was just using the wrong sign. Fixed. In addition, when a straight line coincides with coordinate axis, output appears -0, I do not know whether it is appropriate. This is a matter of the floating-point operations. I don't think it's on scope to worry about that. With the above fix, the cases you pointed out go away anyway. Negative value appeared when use - to calculate the distance between two parallel lines. Fixed with fabs(). New patch included. Still wondering whether to use a A,B,C-based output format per Tom's comment. From d765f8a515a6f3de0e68e68962e5109945260b8c Mon Sep 17 00:00:00 2001 From: Peter Eisentraut pete...@gmx.net Date: Fri, 14 Jun 2013 20:23:15 -0400 Subject: [PATCH] Revive line type Complete the implementations of line_in, line_out, line_recv, line_send. Remove comments and error messages about the line type not being implemented. Add regression tests for existing line operators and functions. Reviewed-by: rui hua 365507506...@gmail.com --- doc/src/sgml/datatype.sgml | 34 +++- doc/src/sgml/func.sgml |6 + src/backend/utils/adt/geo_ops.c| 112 +--- src/include/catalog/pg_type.h |3 +- src/include/utils/geo_decls.h |7 - src/test/regress/expected/geometry.out |3 - src/test/regress/expected/line.out | 267 src/test/regress/expected/sanity_check.out |3 +- src/test/regress/output/misc.source|3 +- src/test/regress/parallel_schedule |2 +- src/test/regress/serial_schedule |1 + src/test/regress/sql/geometry.sql |4 - src/test/regress/sql/line.sql | 86 + 13 files changed, 443 insertions(+), 88 deletions(-) create mode 100644 src/test/regress/expected/line.out create mode 100644 src/test/regress/sql/line.sql diff --git a/doc/src/sgml/datatype.sgml b/doc/src/sgml/datatype.sgml index f73e6b2..ecbbdd8 100644 --- a/doc/src/sgml/datatype.sgml +++ b/doc/src/sgml/datatype.sgml @@ -3066,7 +3066,7 @@ titleGeometric Types/title row entrytypeline/type/entry entry32 bytes/entry -entryInfinite line (not fully implemented)/entry +entryInfinite line/entry entry((x1,y1),(x2,y2))/entry /row row @@ -3142,6 +3142,38 @@ titlePoints/title /sect2 sect2 +titleLines/title + +indexterm + primaryline/primary +/indexterm + +para + Lines (typeline/type) are specified by pairs of points. + Values of type typeline/type are specified using any of the following + syntaxes: + +synopsis +[ ( replaceablex1/replaceable , replaceabley1/replaceable ) , ( replaceablex2/replaceable , replaceabley2/replaceable ) ] +( ( replaceablex1/replaceable , replaceabley1/replaceable ) , ( replaceablex2/replaceable , replaceabley2/replaceable ) ) + ( replaceablex1/replaceable , replaceabley1/replaceable ) , ( replaceablex2/replaceable , replaceabley2/replaceable ) +replaceablex1/replaceable , replaceabley1/replaceable , replaceablex2/replaceable , replaceabley2/replaceable +/synopsis + + where + literal(replaceablex1/replaceable,replaceabley1/replaceable)/literal + and + literal(replaceablex2/replaceable,replaceabley2/replaceable)/literal + are two (different) points on the line. +/para + +para + Lines are output using the first syntax. The points used in the output + are not necessarily the points used on input. +/para + /sect2 + + sect2 titleLine Segments/title indexterm diff --git a/doc/src/sgml/func.sgml b/doc/src/sgml/func.sgml index 8cf5385..fd0e4a3 100644 --- a/doc/src/sgml/func.sgml +++ b/doc/src/sgml/func.sgml @@ -8070,6 +8070,12 @@ titleGeometric Type Conversion Functions/title entryliteralcircle(polygon '((0,0),(1,1),(2,0))')/literal/entry /row row +entryliteralfunctionline(typepoint/type, typepoint/type)/function/literal/entry +entrytypeline/type/entry +entrypoints to line/entry +entryliterallseg(point '(-1,0)', point '(1,0)')/literal/entry + /row + row entry indexterm primarylseg/primary diff --git a/src/backend/utils/adt/geo_ops.c b/src/backend/utils/adt/geo_ops.c index ad18cf0..e668a21 100644 --- a/src/backend/utils/adt/geo_ops.c +++
[HACKERS] [PATCH] Revive line type
Hi, Test results are as follows: Contents Purpose This patch is for finishing the line type and related functions that not done yet but listed in catalogs and documentation. There are no other new features added in this patch. The regression test cases which included in this patch, Documentation are also updated. Run Regression tests are all succeed, but several problems have be found while ding some simple test. The updated document said that the points used in the output are not necessarily the points used on input. I understand that as long as they are marked on the same line. But when the input data represents a horizontal or vertical line, the output is not exactly the same line. It is another line parallel to it. For example: postgres=# select line('1,3,2,3'); line - [(0,-3),(1,-3)] (1 row) postgres=# select line('1,3,1,6'); line - [(-1,0),(-1,1)] (1 row) In addition, when a straight line coincides with coordinate axis, output appears -0, I do not know whether it is appropriate. postgres=# select line('0,1,0,5'); line - [(-0,0),(-0,1)] (1 row) Negative value appeared when use - to calculate the distance between two parallel lines. postgres=# select line('1,1,2,1') - line('-1,-1,-2,-1') ; ?column? -- -2 postgres=# select lseg('1,1,2,1') - line('-1,-1,-2,-1') ; ?column? -- -2 (1 row) The same situation occurs in distance calculation between point and a straight line. postgres=# select point('-1,1') - line('-3,0,-4,0') ; ?column? -- -1 (1 row) Should the distance be positive numbers? Other functions seem work properly. Performance == Because these functions is first implemented. So there is no relatively comparison for the performance. Conclusion This patch lets line type worked. But there are some bugs. Additionally, function close_sl not implemented. With Regards, Rui
Re: [HACKERS] [PATCH] Revive line type
Peter Eisentraut pete...@gmx.net writes: Complete the implementations of line_in, line_out, line_recv, line_send. Remove comments and error messages about the line type not being implemented. Add regression tests for existing line operators and functions. --- This just revives existing functionality, doesn't add anything new. One thing that the original code did not settle was how to convert a line in form Ax+By+C=0 to the two-points output form. Obviously, you can just pick to random points on the line, but I wonder whether there is a more standard solution. ISTM printing a line as two points is an unfortunate representational choice altogether. If the internal form is Ax+By+C=0, shouldn't the text form expose A,B,C directly? What you've got here necessarily suffers a lot of roundoff error during I/O conversions, and so it seems pretty likely to fail dump-and-reload tests. regards, tom lane -- Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
[HACKERS] [PATCH] Revive line type
Complete the implementations of line_in, line_out, line_recv, line_send. Remove comments and error messages about the line type not being implemented. Add regression tests for existing line operators and functions. --- This just revives existing functionality, doesn't add anything new. One thing that the original code did not settle was how to convert a line in form Ax+By+C=0 to the two-points output form. Obviously, you can just pick to random points on the line, but I wonder whether there is a more standard solution. doc/src/sgml/datatype.sgml | 34 +++- doc/src/sgml/func.sgml |6 + src/backend/utils/adt/geo_ops.c| 108 + src/include/catalog/pg_type.h |3 +- src/include/utils/geo_decls.h |7 - src/test/regress/expected/geometry.out |3 - src/test/regress/expected/line.out | 243 src/test/regress/expected/sanity_check.out |3 +- src/test/regress/output/misc.source|3 +- src/test/regress/parallel_schedule |2 +- src/test/regress/serial_schedule |1 + src/test/regress/sql/geometry.sql |4 - src/test/regress/sql/line.sql | 77 + 13 files changed, 408 insertions(+), 86 deletions(-) create mode 100644 src/test/regress/expected/line.out create mode 100644 src/test/regress/sql/line.sql diff --git a/doc/src/sgml/datatype.sgml b/doc/src/sgml/datatype.sgml index f73e6b2..ecbbdd8 100644 --- a/doc/src/sgml/datatype.sgml +++ b/doc/src/sgml/datatype.sgml @@ -3066,7 +3066,7 @@ titleGeometric Types/title row entrytypeline/type/entry entry32 bytes/entry -entryInfinite line (not fully implemented)/entry +entryInfinite line/entry entry((x1,y1),(x2,y2))/entry /row row @@ -3142,6 +3142,38 @@ titlePoints/title /sect2 sect2 +titleLines/title + +indexterm + primaryline/primary +/indexterm + +para + Lines (typeline/type) are specified by pairs of points. + Values of type typeline/type are specified using any of the following + syntaxes: + +synopsis +[ ( replaceablex1/replaceable , replaceabley1/replaceable ) , ( replaceablex2/replaceable , replaceabley2/replaceable ) ] +( ( replaceablex1/replaceable , replaceabley1/replaceable ) , ( replaceablex2/replaceable , replaceabley2/replaceable ) ) + ( replaceablex1/replaceable , replaceabley1/replaceable ) , ( replaceablex2/replaceable , replaceabley2/replaceable ) +replaceablex1/replaceable , replaceabley1/replaceable , replaceablex2/replaceable , replaceabley2/replaceable +/synopsis + + where + literal(replaceablex1/replaceable,replaceabley1/replaceable)/literal + and + literal(replaceablex2/replaceable,replaceabley2/replaceable)/literal + are two (different) points on the line. +/para + +para + Lines are output using the first syntax. The points used in the output + are not necessarily the points used on input. +/para + /sect2 + + sect2 titleLine Segments/title indexterm diff --git a/doc/src/sgml/func.sgml b/doc/src/sgml/func.sgml index 4c5af4b..835a189 100644 --- a/doc/src/sgml/func.sgml +++ b/doc/src/sgml/func.sgml @@ -8070,6 +8070,12 @@ titleGeometric Type Conversion Functions/title entryliteralcircle(polygon '((0,0),(1,1),(2,0))')/literal/entry /row row +entryliteralfunctionline(typepoint/type, typepoint/type)/function/literal/entry +entrytypeline/type/entry +entrypoints to line/entry +entryliterallseg(point '(-1,0)', point '(1,0)')/literal/entry + /row + row entry indexterm primarylseg/primary diff --git a/src/backend/utils/adt/geo_ops.c b/src/backend/utils/adt/geo_ops.c index ad18cf0..61a1900 100644 --- a/src/backend/utils/adt/geo_ops.c +++ b/src/backend/utils/adt/geo_ops.c @@ -933,13 +933,8 @@ Datum line_in(PG_FUNCTION_ARGS) { -#ifdef ENABLE_LINE_TYPE char *str = PG_GETARG_CSTRING(0); -#endif LINE *line; - -#ifdef ENABLE_LINE_TYPE - /* when fixed, modify not implemented, catalog/pg_type.h and SGML */ LSEGlseg; int isopen; char *s; @@ -950,15 +945,13 @@ (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION), errmsg(invalid input syntax for type line: \%s\, str))); + if (FPeq(lseg.p[0].x, lseg.p[1].x) FPeq(lseg.p[0].y, lseg.p[1].y)) + ereport(ERROR, + (errcode(ERRCODE_INVALID_TEXT_REPRESENTATION), +errmsg(invalid line specification: must be two distinct points))); + line = (LINE *) palloc(sizeof(LINE)); line_construct_pts(line, lseg.p[0], lseg.p[1]); -#else - ereport(ERROR, -