Commit d015d08b authored by Tom Lane's avatar Tom Lane

Rename default text search parser's "uri" token type to "url_path",

per recommendation from Alvaro.  This doesn't force initdb since the
numeric token type in the catalogs doesn't change; but note that
the expected regression test output changed.
parent 834ddc62
<!-- $PostgreSQL: pgsql/doc/src/sgml/textsearch.sgml,v 1.27 2007/10/27 00:19:45 tgl Exp $ --> <!-- $PostgreSQL: pgsql/doc/src/sgml/textsearch.sgml,v 1.28 2007/10/27 16:01:08 tgl Exp $ -->
<chapter id="textsearch"> <chapter id="textsearch">
<title id="textsearch-title">Full Text Search</title> <title id="textsearch-title">Full Text Search</title>
...@@ -1815,8 +1815,8 @@ LIMIT 10; ...@@ -1815,8 +1815,8 @@ LIMIT 10;
<entry><literal>example.com</literal></entry> <entry><literal>example.com</literal></entry>
</row> </row>
<row> <row>
<entry><literal>uri</></entry> <entry><literal>url_path</></entry>
<entry>URI</entry> <entry>URL path</entry>
<entry><literal>/stuff/index.html</literal>, in the context of a URL</entry> <entry><literal>/stuff/index.html</literal>, in the context of a URL</entry>
</row> </row>
<row> <row>
...@@ -1907,7 +1907,7 @@ SELECT alias, description, token FROM ts_debug('http://example.com/stuff/index.h ...@@ -1907,7 +1907,7 @@ SELECT alias, description, token FROM ts_debug('http://example.com/stuff/index.h
protocol | Protocol head | http:// protocol | Protocol head | http://
url | URL | example.com/stuff/index.html url | URL | example.com/stuff/index.html
host | Host | example.com host | Host | example.com
uri | URI | /stuff/index.html url_path | URL path | /stuff/index.html
</programlisting> </programlisting>
</para> </para>
...@@ -2632,7 +2632,7 @@ ALTER TEXT SEARCH CONFIGURATION pg ...@@ -2632,7 +2632,7 @@ ALTER TEXT SEARCH CONFIGURATION pg
<programlisting> <programlisting>
ALTER TEXT SEARCH CONFIGURATION pg ALTER TEXT SEARCH CONFIGURATION pg
DROP MAPPING FOR email, url, sfloat, uri, float; DROP MAPPING FOR email, url, url_path, sfloat, float;
</programlisting> </programlisting>
</para> </para>
...@@ -2939,7 +2939,7 @@ SELECT * FROM ts_token_type('default'); ...@@ -2939,7 +2939,7 @@ SELECT * FROM ts_token_type('default');
15 | numhword | Hyphenated word, letters and digits 15 | numhword | Hyphenated word, letters and digits
16 | asciihword | Hyphenated word, all ASCII 16 | asciihword | Hyphenated word, all ASCII
17 | hword | Hyphenated word, all letters 17 | hword | Hyphenated word, all letters
18 | uri | URI 18 | url_path | URL path
19 | file | File or path name 19 | file | File or path name
20 | float | Decimal notation 20 | float | Decimal notation
21 | int | Signed integer 21 | int | Signed integer
...@@ -3308,8 +3308,8 @@ Parser: "pg_catalog.default" ...@@ -3308,8 +3308,8 @@ Parser: "pg_catalog.default"
numword | simple numword | simple
sfloat | simple sfloat | simple
uint | simple uint | simple
uri | simple
url | simple url | simple
url_path | simple
version | simple version | simple
word | russian_stem word | russian_stem
</programlisting> </programlisting>
...@@ -3398,8 +3398,8 @@ Parser: "pg_catalog.default" ...@@ -3398,8 +3398,8 @@ Parser: "pg_catalog.default"
sfloat | Scientific notation sfloat | Scientific notation
tag | HTML tag tag | HTML tag
uint | Unsigned integer uint | Unsigned integer
uri | URI
url | URL url | URL
url_path | URL path
version | Version number version | Version number
word | Word, all letters word | Word, all letters
(23 rows) (23 rows)
......
-- $PostgreSQL: pgsql/src/backend/snowball/snowball.sql.in,v 1.5 2007/10/23 20:46:12 tgl Exp $$ -- $PostgreSQL: pgsql/src/backend/snowball/snowball.sql.in,v 1.6 2007/10/27 16:01:08 tgl Exp $$
-- text search configuration for _LANGNAME_ language -- text search configuration for _LANGNAME_ language
CREATE TEXT SEARCH DICTIONARY _DICTNAME_ CREATE TEXT SEARCH DICTIONARY _DICTNAME_
...@@ -12,8 +12,9 @@ CREATE TEXT SEARCH CONFIGURATION _CFGNAME_ ...@@ -12,8 +12,9 @@ CREATE TEXT SEARCH CONFIGURATION _CFGNAME_
COMMENT ON TEXT SEARCH CONFIGURATION _CFGNAME_ IS 'configuration for _LANGNAME_ language'; COMMENT ON TEXT SEARCH CONFIGURATION _CFGNAME_ IS 'configuration for _LANGNAME_ language';
ALTER TEXT SEARCH CONFIGURATION _CFGNAME_ ADD MAPPING ALTER TEXT SEARCH CONFIGURATION _CFGNAME_ ADD MAPPING
FOR email, url, host, sfloat, version, uri, file, float, int, uint, FOR email, url, url_path, host, file, version,
numword, hword_numpart, numhword sfloat, float, int, uint,
numword, hword_numpart, numhword
WITH simple; WITH simple;
ALTER TEXT SEARCH CONFIGURATION _CFGNAME_ ADD MAPPING ALTER TEXT SEARCH CONFIGURATION _CFGNAME_ ADD MAPPING
......
...@@ -7,7 +7,7 @@ ...@@ -7,7 +7,7 @@
* *
* *
* IDENTIFICATION * IDENTIFICATION
* $PostgreSQL: pgsql/src/backend/tsearch/wparser_def.c,v 1.4 2007/10/23 20:46:12 tgl Exp $ * $PostgreSQL: pgsql/src/backend/tsearch/wparser_def.c,v 1.5 2007/10/27 16:01:08 tgl Exp $
* *
*------------------------------------------------------------------------- *-------------------------------------------------------------------------
*/ */
...@@ -41,7 +41,7 @@ ...@@ -41,7 +41,7 @@
#define NUMHWORD 15 #define NUMHWORD 15
#define ASCIIHWORD 16 #define ASCIIHWORD 16
#define HWORD 17 #define HWORD 17
#define URI 18 #define URLPATH 18
#define FILEPATH 19 #define FILEPATH 19
#define DECIMAL 20 #define DECIMAL 20
#define SIGNEDINT 21 #define SIGNEDINT 21
...@@ -69,7 +69,7 @@ static const char * const tok_alias[] = { ...@@ -69,7 +69,7 @@ static const char * const tok_alias[] = {
"numhword", "numhword",
"asciihword", "asciihword",
"hword", "hword",
"uri", "url_path",
"file", "file",
"float", "float",
"int", "int",
...@@ -96,7 +96,7 @@ static const char * const lex_descr[] = { ...@@ -96,7 +96,7 @@ static const char * const lex_descr[] = {
"Hyphenated word, letters and digits", "Hyphenated word, letters and digits",
"Hyphenated word, all ASCII", "Hyphenated word, all ASCII",
"Hyphenated word, all letters", "Hyphenated word, all letters",
"URI", "URL path",
"File or path name", "File or path name",
"Decimal notation", "Decimal notation",
"Signed integer", "Signed integer",
...@@ -164,9 +164,9 @@ typedef enum ...@@ -164,9 +164,9 @@ typedef enum
TPS_InPathSecond, TPS_InPathSecond,
TPS_InFile, TPS_InFile,
TPS_InFileNext, TPS_InFileNext,
TPS_InURIFirst, TPS_InURLPathFirst,
TPS_InURIStart, TPS_InURLPathStart,
TPS_InURI, TPS_InURLPath,
TPS_InFURL, TPS_InFURL,
TPS_InProtocolFirst, TPS_InProtocolFirst,
TPS_InProtocolSecond, TPS_InProtocolSecond,
...@@ -624,7 +624,7 @@ p_ishost(TParser * prs) ...@@ -624,7 +624,7 @@ p_ishost(TParser * prs)
} }
static int static int
p_isURI(TParser * prs) p_isURLPath(TParser * prs)
{ {
TParser *tmpprs = TParserInit(prs->str + prs->state->posbyte, prs->lenstr - prs->state->posbyte); TParser *tmpprs = TParserInit(prs->str + prs->state->posbyte, prs->lenstr - prs->state->posbyte);
int res = 0; int res = 0;
...@@ -632,7 +632,7 @@ p_isURI(TParser * prs) ...@@ -632,7 +632,7 @@ p_isURI(TParser * prs)
tmpprs->state = newTParserPosition(tmpprs->state); tmpprs->state = newTParserPosition(tmpprs->state);
tmpprs->state->state = TPS_InFileFirst; tmpprs->state->state = TPS_InFileFirst;
if (TParserGet(tmpprs) && (tmpprs->type == URI || tmpprs->type == FILEPATH)) if (TParserGet(tmpprs) && (tmpprs->type == URLPATH || tmpprs->type == FILEPATH))
{ {
prs->state->posbyte += tmpprs->lenbytelexeme; prs->state->posbyte += tmpprs->lenbytelexeme;
prs->state->poschar += tmpprs->lencharlexeme; prs->state->poschar += tmpprs->lencharlexeme;
...@@ -995,7 +995,7 @@ static TParserStateActionItem actionTPS_InHostDomain[] = { ...@@ -995,7 +995,7 @@ static TParserStateActionItem actionTPS_InHostDomain[] = {
{p_iseqC, '.', A_PUSH, TPS_InHostFirstDomain, 0, NULL}, {p_iseqC, '.', A_PUSH, TPS_InHostFirstDomain, 0, NULL},
{p_iseqC, '@', A_PUSH, TPS_InEmail, 0, NULL}, {p_iseqC, '@', A_PUSH, TPS_InEmail, 0, NULL},
{p_isdigit, 0, A_POP, TPS_Null, 0, NULL}, {p_isdigit, 0, A_POP, TPS_Null, 0, NULL},
{p_isstophost, 0, A_BINGO | A_CLRALL, TPS_InURIStart, HOST, NULL}, {p_isstophost, 0, A_BINGO | A_CLRALL, TPS_InURLPathStart, HOST, NULL},
{p_iseqC, '/', A_PUSH, TPS_InFURL, 0, NULL}, {p_iseqC, '/', A_PUSH, TPS_InFURL, 0, NULL},
{NULL, 0, A_BINGO | A_CLRALL, TPS_Base, HOST, NULL} {NULL, 0, A_BINGO | A_CLRALL, TPS_Base, HOST, NULL}
}; };
...@@ -1009,7 +1009,7 @@ static TParserStateActionItem actionTPS_InPortFirst[] = { ...@@ -1009,7 +1009,7 @@ static TParserStateActionItem actionTPS_InPortFirst[] = {
static TParserStateActionItem actionTPS_InPort[] = { static TParserStateActionItem actionTPS_InPort[] = {
{p_isEOF, 0, A_BINGO | A_CLRALL, TPS_Base, HOST, NULL}, {p_isEOF, 0, A_BINGO | A_CLRALL, TPS_Base, HOST, NULL},
{p_isdigit, 0, A_NEXT, TPS_InPort, 0, NULL}, {p_isdigit, 0, A_NEXT, TPS_InPort, 0, NULL},
{p_isstophost, 0, A_BINGO | A_CLRALL, TPS_InURIStart, HOST, NULL}, {p_isstophost, 0, A_BINGO | A_CLRALL, TPS_InURLPathStart, HOST, NULL},
{p_iseqC, '/', A_PUSH, TPS_InFURL, 0, NULL}, {p_iseqC, '/', A_PUSH, TPS_InFURL, 0, NULL},
{NULL, 0, A_BINGO | A_CLRALL, TPS_Base, HOST, NULL} {NULL, 0, A_BINGO | A_CLRALL, TPS_Base, HOST, NULL}
}; };
...@@ -1042,7 +1042,7 @@ static TParserStateActionItem actionTPS_InFileFirst[] = { ...@@ -1042,7 +1042,7 @@ static TParserStateActionItem actionTPS_InFileFirst[] = {
{p_isdigit, 0, A_NEXT, TPS_InFile, 0, NULL}, {p_isdigit, 0, A_NEXT, TPS_InFile, 0, NULL},
{p_iseqC, '.', A_NEXT, TPS_InPathFirst, 0, NULL}, {p_iseqC, '.', A_NEXT, TPS_InPathFirst, 0, NULL},
{p_iseqC, '_', A_NEXT, TPS_InFile, 0, NULL}, {p_iseqC, '_', A_NEXT, TPS_InFile, 0, NULL},
{p_iseqC, '?', A_PUSH, TPS_InURIFirst, 0, NULL}, {p_iseqC, '?', A_PUSH, TPS_InURLPathFirst, 0, NULL},
{p_iseqC, '~', A_PUSH, TPS_InFileTwiddle, 0, NULL}, {p_iseqC, '~', A_PUSH, TPS_InFileTwiddle, 0, NULL},
{NULL, 0, A_POP, TPS_Null, 0, NULL} {NULL, 0, A_POP, TPS_Null, 0, NULL}
}; };
...@@ -1089,7 +1089,7 @@ static TParserStateActionItem actionTPS_InFile[] = { ...@@ -1089,7 +1089,7 @@ static TParserStateActionItem actionTPS_InFile[] = {
{p_iseqC, '_', A_NEXT, TPS_InFile, 0, NULL}, {p_iseqC, '_', A_NEXT, TPS_InFile, 0, NULL},
{p_iseqC, '-', A_NEXT, TPS_InFile, 0, NULL}, {p_iseqC, '-', A_NEXT, TPS_InFile, 0, NULL},
{p_iseqC, '/', A_PUSH, TPS_InFileFirst, 0, NULL}, {p_iseqC, '/', A_PUSH, TPS_InFileFirst, 0, NULL},
{p_iseqC, '?', A_PUSH, TPS_InURIFirst, 0, NULL}, {p_iseqC, '?', A_PUSH, TPS_InURLPathFirst, 0, NULL},
{NULL, 0, A_BINGO, TPS_Base, FILEPATH, NULL} {NULL, 0, A_BINGO, TPS_Base, FILEPATH, NULL}
}; };
...@@ -1101,29 +1101,29 @@ static TParserStateActionItem actionTPS_InFileNext[] = { ...@@ -1101,29 +1101,29 @@ static TParserStateActionItem actionTPS_InFileNext[] = {
{NULL, 0, A_POP, TPS_Null, 0, NULL} {NULL, 0, A_POP, TPS_Null, 0, NULL}
}; };
static TParserStateActionItem actionTPS_InURIFirst[] = { static TParserStateActionItem actionTPS_InURLPathFirst[] = {
{p_isEOF, 0, A_POP, TPS_Null, 0, NULL}, {p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
{p_iseqC, '"', A_POP, TPS_Null, 0, NULL}, {p_iseqC, '"', A_POP, TPS_Null, 0, NULL},
{p_iseqC, '\'', A_POP, TPS_Null, 0, NULL}, {p_iseqC, '\'', A_POP, TPS_Null, 0, NULL},
{p_isnotspace, 0, A_CLEAR, TPS_InURI, 0, NULL}, {p_isnotspace, 0, A_CLEAR, TPS_InURLPath, 0, NULL},
{NULL, 0, A_POP, TPS_Null, 0, NULL}, {NULL, 0, A_POP, TPS_Null, 0, NULL},
}; };
static TParserStateActionItem actionTPS_InURIStart[] = { static TParserStateActionItem actionTPS_InURLPathStart[] = {
{NULL, 0, A_NEXT, TPS_InURI, 0, NULL} {NULL, 0, A_NEXT, TPS_InURLPath, 0, NULL}
}; };
static TParserStateActionItem actionTPS_InURI[] = { static TParserStateActionItem actionTPS_InURLPath[] = {
{p_isEOF, 0, A_BINGO, TPS_Base, URI, NULL}, {p_isEOF, 0, A_BINGO, TPS_Base, URLPATH, NULL},
{p_iseqC, '"', A_BINGO, TPS_Base, URI, NULL}, {p_iseqC, '"', A_BINGO, TPS_Base, URLPATH, NULL},
{p_iseqC, '\'', A_BINGO, TPS_Base, URI, NULL}, {p_iseqC, '\'', A_BINGO, TPS_Base, URLPATH, NULL},
{p_isnotspace, 0, A_NEXT, TPS_InURI, 0, NULL}, {p_isnotspace, 0, A_NEXT, TPS_InURLPath, 0, NULL},
{NULL, 0, A_BINGO, TPS_Base, URI, NULL} {NULL, 0, A_BINGO, TPS_Base, URLPATH, NULL}
}; };
static TParserStateActionItem actionTPS_InFURL[] = { static TParserStateActionItem actionTPS_InFURL[] = {
{p_isEOF, 0, A_POP, TPS_Null, 0, NULL}, {p_isEOF, 0, A_POP, TPS_Null, 0, NULL},
{p_isURI, 0, A_BINGO | A_CLRALL, TPS_Base, URL_T, SpecialFURL}, {p_isURLPath, 0, A_BINGO | A_CLRALL, TPS_Base, URL_T, SpecialFURL},
{NULL, 0, A_POP, TPS_Null, 0, NULL} {NULL, 0, A_POP, TPS_Null, 0, NULL}
}; };
...@@ -1344,9 +1344,9 @@ static const TParserStateAction Actions[] = { ...@@ -1344,9 +1344,9 @@ static const TParserStateAction Actions[] = {
{TPS_InPathSecond, actionTPS_InPathSecond}, {TPS_InPathSecond, actionTPS_InPathSecond},
{TPS_InFile, actionTPS_InFile}, {TPS_InFile, actionTPS_InFile},
{TPS_InFileNext, actionTPS_InFileNext}, {TPS_InFileNext, actionTPS_InFileNext},
{TPS_InURIFirst, actionTPS_InURIFirst}, {TPS_InURLPathFirst, actionTPS_InURLPathFirst},
{TPS_InURIStart, actionTPS_InURIStart}, {TPS_InURLPathStart, actionTPS_InURLPathStart},
{TPS_InURI, actionTPS_InURI}, {TPS_InURLPath, actionTPS_InURLPath},
{TPS_InFURL, actionTPS_InFURL}, {TPS_InFURL, actionTPS_InFURL},
{TPS_InProtocolFirst, actionTPS_InProtocolFirst}, {TPS_InProtocolFirst, actionTPS_InProtocolFirst},
{TPS_InProtocolSecond, actionTPS_InProtocolSecond}, {TPS_InProtocolSecond, actionTPS_InProtocolSecond},
......
...@@ -227,7 +227,7 @@ SELECT * FROM ts_token_type('default'); ...@@ -227,7 +227,7 @@ SELECT * FROM ts_token_type('default');
15 | numhword | Hyphenated word, letters and digits 15 | numhword | Hyphenated word, letters and digits
16 | asciihword | Hyphenated word, all ASCII 16 | asciihword | Hyphenated word, all ASCII
17 | hword | Hyphenated word, all letters 17 | hword | Hyphenated word, all letters
18 | uri | URI 18 | url_path | URL path
19 | file | File or path name 19 | file | File or path name
20 | float | Decimal notation 20 | float | Decimal notation
21 | int | Signed integer 21 | int | Signed integer
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment