Build all Flex files standalone

The proposed Meson build system will need a way to ignore certain
generated files in order to coexist with the autoconf build system,
and C files generated by Flex which are #include'd into .y files make
this more difficult. In similar vein to 72b1e3a21, arrange for all Flex
C files to compile to their own .o targets.

Reviewed by Andres Freund

Discussion: https://www.postgresql.org/message-id/20220810171935.7k5zgnjwqzalzmtm%40awork3.anarazel.de
Discussion: https://www.postgresql.org/message-id/CAFBsxsF8Gc2StS3haXofshHCzqNMRXiSxvQEYGwnFsTmsdwNeg@mail.gmail.com

contrib/cube/.gitignore

@@ -1,3 +1,4 @@
+/cubeparse.h
 /cubeparse.c
 /cubescan.c
 # Generated subdirectories

contrib/cube/Makefile

@@ -4,7 +4,8 @@ MODULE_big = cube
 OBJS = \
 	$(WIN32RES) \
 	cube.o \
-	cubeparse.o
+	cubeparse.o \
+	cubescan.o
 
 EXTENSION = cube
 DATA = cube--1.2.sql cube--1.2--1.3.sql cube--1.3--1.4.sql cube--1.4--1.5.sql \
@@ -15,8 +16,6 @@ HEADERS = cubedata.h
 
 REGRESS = cube cube_sci
 
-EXTRA_CLEAN = y.tab.c y.tab.h
-
 SHLIB_LINK += $(filter -lm, $(LIBS))
 
 ifdef USE_PGXS
@@ -30,11 +29,16 @@ include $(top_builddir)/src/Makefile.global
 include $(top_srcdir)/contrib/contrib-global.mk
 endif
 
+# See notes in src/backend/parser/Makefile about the following two rules
+cubeparse.h: cubeparse.c
+	touch $@
+
+cubeparse.c: BISONFLAGS += -d
 
-# cubescan is compiled as part of cubeparse
-cubeparse.o: cubescan.c
+# Force these dependencies to be known even without dependency info built:
+cubeparse.o cubescan.o: cubeparse.h
 
 distprep: cubeparse.c cubescan.c
 
 maintainer-clean:
-	rm -f cubeparse.c cubescan.c
+	rm -f cubeparse.h cubeparse.c cubescan.c

contrib/cube/cube.c

@@ -119,11 +119,11 @@ cube_in(PG_FUNCTION_ARGS)
 {
 	char	   *str = PG_GETARG_CSTRING(0);
 	NDBOX	   *result;
+	Size		scanbuflen;
 
-	cube_scanner_init(str);
+	cube_scanner_init(str, &scanbuflen);
 
-	if (cube_yyparse(&result) != 0)
-		cube_yyerror(&result, "cube parser failed");
+	cube_yyparse(&result, scanbuflen);
 
 	cube_scanner_finish();
 

contrib/cube/cubedata.h

@@ -61,9 +61,9 @@ typedef struct NDBOX
 
 /* in cubescan.l */
 extern int	cube_yylex(void);
-extern void cube_yyerror(NDBOX **result, const char *message) pg_attribute_noreturn();
-extern void cube_scanner_init(const char *str);
+extern void cube_yyerror(NDBOX **result, Size scanbuflen, const char *message) pg_attribute_noreturn();
+extern void cube_scanner_init(const char *str, Size *scanbuflen);
 extern void cube_scanner_finish(void);
 
 /* in cubeparse.y */
-extern int	cube_yyparse(NDBOX **result);
+extern int	cube_yyparse(NDBOX **result, Size scanbuflen);

contrib/cube/cubeparse.y

@@ -23,9 +23,6 @@
 #define YYMALLOC palloc
 #define YYFREE   pfree
 
-static char *scanbuf;
-static int	scanbuflen;
-
 static int item_count(const char *s, char delim);
 static NDBOX *write_box(int dim, char *str1, char *str2);
 static NDBOX *write_point_as_box(int dim, char *str);
@@ -34,6 +31,7 @@ static NDBOX *write_point_as_box(int dim, char *str);
 
 /* BISON Declarations */
 %parse-param {NDBOX **result}
+%parse-param {Size scanbuflen}
 %expect 0
 %name-prefix="cube_yy"
 
@@ -265,5 +263,3 @@ write_point_as_box(int dim, char *str)
 
 	return bp;
 }
-
-#include "cubescan.c"

contrib/cube/cubescan.l

@@ -1,9 +1,21 @@
-%{
+%top{
 /*
  * A scanner for EMP-style numeric ranges
  * contrib/cube/cubescan.l
  */
 
+#include "postgres.h"
+
+/*
+ * NB: include cubeparse.h only AFTER defining YYSTYPE (to match cubeparse.y)
+ * and cubedata.h for NDBOX.
+ */
+#include "cubedata.h"
+#define YYSTYPE char *
+#include "cubeparse.h"
+}
+
+%{
 /* LCOV_EXCL_START */
 
 /* No reason to constrain amount of data slurped */
@@ -21,9 +33,7 @@ fprintf_to_ereport(const char *fmt, const char *msg)
 
 /* Handles to the buffer that the lexer uses internally */
 static YY_BUFFER_STATE scanbufhandle;
-/* this is now declared in cubeparse.y: */
-/* static char *scanbuf; */
-/* static int	scanbuflen; */
+static char *scanbuf;
 %}
 
 %option 8bit
@@ -45,14 +55,14 @@ NaN          [nN][aA][nN]
 
 %%
 
-{float}      yylval = yytext; return CUBEFLOAT;
-{infinity}   yylval = yytext; return CUBEFLOAT;
-{NaN}        yylval = yytext; return CUBEFLOAT;
-\[           yylval = "("; return O_BRACKET;
-\]           yylval = ")"; return C_BRACKET;
-\(           yylval = "("; return O_PAREN;
-\)           yylval = ")"; return C_PAREN;
-\,           yylval = ","; return COMMA;
+{float}      cube_yylval = yytext; return CUBEFLOAT;
+{infinity}   cube_yylval = yytext; return CUBEFLOAT;
+{NaN}        cube_yylval = yytext; return CUBEFLOAT;
+\[           cube_yylval = "("; return O_BRACKET;
+\]           cube_yylval = ")"; return C_BRACKET;
+\(           cube_yylval = "("; return O_PAREN;
+\)           cube_yylval = ")"; return C_PAREN;
+\,           cube_yylval = ","; return COMMA;
 [ \t\n\r\f]+ /* discard spaces */
 .            return yytext[0]; /* alert parser of the garbage */
 
@@ -60,9 +70,9 @@ NaN          [nN][aA][nN]
 
 /* LCOV_EXCL_STOP */
 
-/* result is not used, but Bison expects this signature */
+/* result and scanbuflen are not used, but Bison expects this signature */
 void
-yyerror(NDBOX **result, const char *message)
+cube_yyerror(NDBOX **result, Size scanbuflen, const char *message)
 {
 	if (*yytext == YY_END_OF_BUFFER_CHAR)
 	{
@@ -87,9 +97,9 @@ yyerror(NDBOX **result, const char *message)
  * Called before any actual parsing is done
  */
 void
-cube_scanner_init(const char *str)
+cube_scanner_init(const char *str, Size *scanbuflen)
 {
-	Size	slen = strlen(str);
+	Size		slen = strlen(str);
 
 	/*
 	 * Might be left over after ereport()
@@ -100,7 +110,7 @@ cube_scanner_init(const char *str)
 	/*
 	 * Make a scan buffer with special termination needed by flex.
 	 */
-	scanbuflen = slen;
+	*scanbuflen = slen;
 	scanbuf = palloc(slen + 2);
 	memcpy(scanbuf, str, slen);
 	scanbuf[slen] = scanbuf[slen + 1] = YY_END_OF_BUFFER_CHAR;

contrib/seg/.gitignore

@@ -1,3 +1,4 @@
+/segparse.h
 /segparse.c
 /segscan.c
 # Generated subdirectories

contrib/seg/Makefile

@@ -4,7 +4,8 @@ MODULE_big = seg
 OBJS = \
 	$(WIN32RES) \
 	seg.o \
-	segparse.o
+	segparse.o \
+	segscan.o
 
 EXTENSION = seg
 DATA = seg--1.1.sql seg--1.1--1.2.sql seg--1.2--1.3.sql seg--1.3--1.4.sql \
@@ -29,10 +30,16 @@ include $(top_srcdir)/contrib/contrib-global.mk
 endif
 
 
-# segscan is compiled as part of segparse
-segparse.o: segscan.c
+# See notes in src/backend/parser/Makefile about the following two rules
+segparse.h: segparse.c
+	touch $@
+
+segparse.c: BISONFLAGS += -d
+
+# Force these dependencies to be known even without dependency info built:
+segparse.o segscan.o: segparse.h
 
 distprep: segparse.c segscan.c
 
 maintainer-clean:
-	rm -f segparse.c segscan.c
+	rm -f segparse.h segparse.c segscan.c

contrib/seg/segparse.y

@@ -160,6 +160,3 @@ seg_atof(const char *value)
 	datum = DirectFunctionCall1(float4in, CStringGetDatum(value));
 	return DatumGetFloat4(datum);
 }
-
-
-#include "segscan.c"

contrib/seg/segscan.l

@@ -1,8 +1,18 @@
-%{
+%top{
 /*
  * A scanner for EMP-style numeric ranges
  */
+#include "postgres.h"
+
+/*
+ * NB: include segparse.h only AFTER including segdata.h, because segdata.h
+ * contains the definition for SEG.
+ */
+#include "segdata.h"
+#include "segparse.h"
+}
 
+%{
 /* LCOV_EXCL_START */
 
 /* No reason to constrain amount of data slurped */
@@ -21,7 +31,6 @@ fprintf_to_ereport(const char *fmt, const char *msg)
 /* Handles to the buffer that the lexer uses internally */
 static YY_BUFFER_STATE scanbufhandle;
 static char *scanbuf;
-static int	scanbuflen;
 %}
 
 %option 8bit
@@ -42,12 +51,12 @@ float        ({integer}|{real})([eE]{integer})?
 
 %%
 
-{range}      yylval.text = yytext; return RANGE;
-{plumin}     yylval.text = yytext; return PLUMIN;
-{float}      yylval.text = yytext; return SEGFLOAT;
-\<           yylval.text = "<"; return EXTENSION;
-\>           yylval.text = ">"; return EXTENSION;
-\~           yylval.text = "~"; return EXTENSION;
+{range}      seg_yylval.text = yytext; return RANGE;
+{plumin}     seg_yylval.text = yytext; return PLUMIN;
+{float}      seg_yylval.text = yytext; return SEGFLOAT;
+\<           seg_yylval.text = "<"; return EXTENSION;
+\>           seg_yylval.text = ">"; return EXTENSION;
+\~           seg_yylval.text = "~"; return EXTENSION;
 [ \t\n\r\f]+ /* discard spaces */
 .            return yytext[0]; /* alert parser of the garbage */
 
@@ -56,7 +65,7 @@ float        ({integer}|{real})([eE]{integer})?
 /* LCOV_EXCL_STOP */
 
 void
-yyerror(SEG *result, const char *message)
+seg_yyerror(SEG *result, const char *message)
 {
 	if (*yytext == YY_END_OF_BUFFER_CHAR)
 	{
@@ -94,7 +103,6 @@ seg_scanner_init(const char *str)
 	/*
 	 * Make a scan buffer with special termination needed by flex.
 	 */
-	scanbuflen = slen;
 	scanbuf = palloc(slen + 2);
 	memcpy(scanbuf, str, slen);
 	scanbuf[slen] = scanbuf[slen + 1] = YY_END_OF_BUFFER_CHAR;

src/backend/Makefile

@@ -174,10 +174,10 @@ utils/probes.o: utils/probes.d $(SUBDIROBJS)
 # Be sure that these files get removed by the maintainer-clean target
 distprep:
 	$(MAKE) -C parser	gram.c gram.h scan.c
-	$(MAKE) -C bootstrap	bootparse.c bootscanner.c
+	$(MAKE) -C bootstrap	bootparse.c bootparse.h bootscanner.c
 	$(MAKE) -C catalog	distprep
 	$(MAKE) -C nodes	distprep
-	$(MAKE) -C replication	repl_gram.c repl_scanner.c syncrep_gram.c syncrep_scanner.c
+	$(MAKE) -C replication	repl_gram.c repl_gram.h repl_scanner.c syncrep_gram.c syncrep_gram.h syncrep_scanner.c
 	$(MAKE) -C storage/lmgr	lwlocknames.h lwlocknames.c
 	$(MAKE) -C utils	distprep
 	$(MAKE) -C utils/adt	jsonpath_gram.c jsonpath_scan.c
@@ -292,13 +292,16 @@ maintainer-clean: distclean
 	$(MAKE) -C nodes $@
 	$(MAKE) -C utils $@
 	rm -f bootstrap/bootparse.c \
+	      bootstrap/bootparse.h \
 	      bootstrap/bootscanner.c \
 	      parser/gram.c \
 	      parser/gram.h \
 	      parser/scan.c \
 	      replication/repl_gram.c \
+	      replication/repl_gram.h \
 	      replication/repl_scanner.c \
 	      replication/syncrep_gram.c \
+	      replication/syncrep_gram.h \
 	      replication/syncrep_scanner.c \
 	      storage/lmgr/lwlocknames.c \
 	      storage/lmgr/lwlocknames.h \

src/backend/bootstrap/.gitignore

@@ -1,2 +1,3 @@
+/bootparse.h
 /bootparse.c
 /bootscanner.c

src/backend/bootstrap/Makefile

@@ -14,12 +14,19 @@ override CPPFLAGS := -I. -I$(srcdir) $(CPPFLAGS)
 
 OBJS = \
 	bootparse.o \
+	bootscanner.o \
 	bootstrap.o
 
 include $(top_srcdir)/src/backend/common.mk
 
-# bootscanner is compiled as part of bootparse
-bootparse.o: bootscanner.c
+# See notes in src/backend/parser/Makefile about the following two rules
+bootparse.h: bootparse.c
+	touch $@
+
+bootparse.c: BISONFLAGS += -d
+
+# Force these dependencies to be known even without dependency info built:
+bootparse.o bootscanner.o: bootparse.h
 
 # bootparse.c and bootscanner.c are in the distribution tarball, so
 # they are not cleaned here.

src/backend/bootstrap/bootparse.y

@@ -488,5 +488,3 @@ boot_ident:
 		| XNULL			{ $$ = pstrdup($1); }
 		;
 %%
-
-#include "bootscanner.c"

src/backend/bootstrap/bootscanner.l

@@ -1,4 +1,4 @@
-%{
+%top{
 /*-------------------------------------------------------------------------
  *
  * bootscanner.l
@@ -15,11 +15,17 @@
  */
 #include "postgres.h"
 
+/*
+ * NB: include bootparse.h only AFTER including bootstrap.h, because bootstrap.h
+ * includes node definitions needed for YYSTYPE.
+ */
 #include "bootstrap/bootstrap.h"
+#include "bootparse.h"
 #include "utils/guc.h"
 
-/* Not needed now that this file is compiled as part of bootparse. */
-/* #include "bootparse.h" */
+}
+
+%{
 
 /* LCOV_EXCL_START */
 
@@ -52,7 +58,7 @@ id		[-A-Za-z0-9_]+
 sid		\'([^']|\'\')*\'
 
 /*
- * Keyword tokens return the keyword text (as a constant string) in yylval.kw,
+ * Keyword tokens return the keyword text (as a constant string) in boot_yylval.kw,
  * just in case that's needed because we want to treat the keyword as an
  * unreserved identifier.  Note that _null_ is not treated as a keyword
  * for this purpose; it's the one "reserved word" in the bootstrap syntax.
@@ -60,23 +66,23 @@ sid		\'([^']|\'\')*\'
  * Notice that all the keywords are case-sensitive, and for historical
  * reasons some must be upper case.
  *
- * String tokens return a palloc'd string in yylval.str.
+ * String tokens return a palloc'd string in boot_yylval.str.
  */
 
 %%
 
-open			{ yylval.kw = "open"; return OPEN; }
+open			{ boot_yylval.kw = "open"; return OPEN; }
 
-close			{ yylval.kw = "close"; return XCLOSE; }
+close			{ boot_yylval.kw = "close"; return XCLOSE; }
 
-create			{ yylval.kw = "create"; return XCREATE; }
+create			{ boot_yylval.kw = "create"; return XCREATE; }
 
-OID				{ yylval.kw = "OID"; return OBJ_ID; }
-bootstrap		{ yylval.kw = "bootstrap"; return XBOOTSTRAP; }
-shared_relation	{ yylval.kw = "shared_relation"; return XSHARED_RELATION; }
-rowtype_oid		{ yylval.kw = "rowtype_oid"; return XROWTYPE_OID; }
+OID				{ boot_yylval.kw = "OID"; return OBJ_ID; }
+bootstrap		{ boot_yylval.kw = "bootstrap"; return XBOOTSTRAP; }
+shared_relation	{ boot_yylval.kw = "shared_relation"; return XSHARED_RELATION; }
+rowtype_oid		{ boot_yylval.kw = "rowtype_oid"; return XROWTYPE_OID; }
 
-insert			{ yylval.kw = "insert"; return INSERT_TUPLE; }
+insert			{ boot_yylval.kw = "insert"; return INSERT_TUPLE; }
 
 _null_			{ return NULLVAL; }
 
@@ -90,25 +96,25 @@ _null_			{ return NULLVAL; }
 
 ^\#[^\n]*		;		/* drop everything after "#" for comments */
 
-declare			{ yylval.kw = "declare"; return XDECLARE; }
-build			{ yylval.kw = "build"; return XBUILD; }
-indices			{ yylval.kw = "indices"; return INDICES; }
-unique			{ yylval.kw = "unique"; return UNIQUE; }
-index			{ yylval.kw = "index"; return INDEX; }
-on				{ yylval.kw = "on"; return ON; }
-using			{ yylval.kw = "using"; return USING; }
-toast			{ yylval.kw = "toast"; return XTOAST; }
-FORCE			{ yylval.kw = "FORCE"; return XFORCE; }
-NOT				{ yylval.kw = "NOT"; return XNOT; }
-NULL			{ yylval.kw = "NULL"; return XNULL; }
+declare			{ boot_yylval.kw = "declare"; return XDECLARE; }
+build			{ boot_yylval.kw = "build"; return XBUILD; }
+indices			{ boot_yylval.kw = "indices"; return INDICES; }
+unique			{ boot_yylval.kw = "unique"; return UNIQUE; }
+index			{ boot_yylval.kw = "index"; return INDEX; }
+on				{ boot_yylval.kw = "on"; return ON; }
+using			{ boot_yylval.kw = "using"; return USING; }
+toast			{ boot_yylval.kw = "toast"; return XTOAST; }
+FORCE			{ boot_yylval.kw = "FORCE"; return XFORCE; }
+NOT				{ boot_yylval.kw = "NOT"; return XNOT; }
+NULL			{ boot_yylval.kw = "NULL"; return XNULL; }
 
 {id}			{
-					yylval.str = pstrdup(yytext);
+					boot_yylval.str = pstrdup(yytext);
 					return ID;
 				}
 {sid}			{
 					/* strip quotes and escapes */
-					yylval.str = DeescapeQuotedString(yytext);
+					boot_yylval.str = DeescapeQuotedString(yytext);
 					return ID;
 				}
 
@@ -121,7 +127,7 @@ NULL			{ yylval.kw = "NULL"; return XNULL; }
 /* LCOV_EXCL_STOP */
 
 void
-yyerror(const char *message)
+boot_yyerror(const char *message)
 {
 	elog(ERROR, "%s at line %d", message, yyline);
 }

src/backend/replication/.gitignore

@@ -1,4 +1,6 @@
+/repl_gram.h
 /repl_gram.c
 /repl_scanner.c
+/syncrep_gram.h
 /syncrep_gram.c
 /syncrep_scanner.c

src/backend/replication/Makefile

@@ -16,10 +16,12 @@ override CPPFLAGS := -I. -I$(srcdir) $(CPPFLAGS)
 
 OBJS = \
 	repl_gram.o \
+	repl_scanner.o \
 	slot.o \
 	slotfuncs.o \
 	syncrep.o \
 	syncrep_gram.o \
+	syncrep_scanner.o \
 	walreceiver.o \
 	walreceiverfuncs.o \
 	walsender.o
@@ -28,11 +30,23 @@ SUBDIRS = logical
 
 include $(top_srcdir)/src/backend/common.mk
 
-# repl_scanner is compiled as part of repl_gram
-repl_gram.o: repl_scanner.c
+# See notes in src/backend/parser/Makefile about the following two rules
+repl_gram.h: repl_gram.c
+	touch $@
 
-# syncrep_scanner is compiled as part of syncrep_gram
-syncrep_gram.o: syncrep_scanner.c
+repl_gram.c: BISONFLAGS += -d
+
+# Force these dependencies to be known even without dependency info built:
+repl_gram.o repl_scanner.o: repl_gram.h
+
+# See notes in src/backend/parser/Makefile about the following two rules
+syncrep_gram.h: syncrep_gram.c
+	touch $@
+
+syncrep_gram.c: BISONFLAGS += -d
+
+# Force these dependencies to be known even without dependency info built:
+syncrep_gram.o syncrep_scanner.o: syncrep_gram.h
 
 # repl_gram.c, repl_scanner.c, syncrep_gram.c and syncrep_scanner.c
 # are in the distribution tarball, so they are not cleaned here.

src/backend/replication/repl_gram.y

@@ -416,5 +416,3 @@ ident_or_keyword:
 		;
 
 %%
-
-#include "repl_scanner.c"

src/backend/replication/repl_scanner.l

@@ -1,4 +1,4 @@
-%{
+%top{
 /*-------------------------------------------------------------------------
  *
  * repl_scanner.l
@@ -18,6 +18,15 @@
 #include "utils/builtins.h"
 #include "parser/scansup.h"
 
+/*
+ * NB: include repl_gram.h only AFTER including walsender_private.h, because
+ * walsender_private includes headers that define XLogRecPtr.
+ */
+#include "replication/walsender_private.h"
+#include "repl_gram.h"
+}
+
+%{
 /* Avoid exit() on fatal scanner errors (a bit ugly -- see yy_fatal_error) */
 #undef fprintf
 #define fprintf(file, fmt, msg)  fprintf_to_ereport(fmt, msg)
@@ -130,7 +139,7 @@ WAIT				{ return K_WAIT; }
 {space}+		{ /* do nothing */ }
 
 {digit}+		{
-					yylval.uintval = strtoul(yytext, NULL, 10);
+					replication_yylval.uintval = strtoul(yytext, NULL, 10);
 					return UCONST;
 				}
 
@@ -138,8 +147,8 @@ WAIT				{ return K_WAIT; }
 					uint32	hi,
 							lo;
 					if (sscanf(yytext, "%X/%X", &hi, &lo) != 2)
-						yyerror("invalid streaming start location");
-					yylval.recptr = ((uint64) hi) << 32 | lo;
+						replication_yyerror("invalid streaming start location");
+					replication_yylval.recptr = ((uint64) hi) << 32 | lo;
 					return RECPTR;
 				}
 
@@ -151,7 +160,7 @@ WAIT				{ return K_WAIT; }
 <xq>{quotestop}	{
 					yyless(1);
 					BEGIN(INITIAL);
-					yylval.str = litbufdup();
+					replication_yylval.str = litbufdup();
 					return SCONST;
 				}
 
@@ -173,9 +182,9 @@ WAIT				{ return K_WAIT; }
 
 					yyless(1);
 					BEGIN(INITIAL);
-					yylval.str = litbufdup();
-					len = strlen(yylval.str);
-					truncate_identifier(yylval.str, len, true);
+					replication_yylval.str = litbufdup();
+					len = strlen(replication_yylval.str);
+					truncate_identifier(replication_yylval.str, len, true);
 					return IDENT;
 				}
 
@@ -186,7 +195,7 @@ WAIT				{ return K_WAIT; }
 {identifier}	{
 					int			len = strlen(yytext);
 
-					yylval.str = downcase_truncate_identifier(yytext, len, true);
+					replication_yylval.str = downcase_truncate_identifier(yytext, len, true);
 					return IDENT;
 				}
 
@@ -195,7 +204,7 @@ WAIT				{ return K_WAIT; }
 					return yytext[0];
 				}
 
-<xq,xd><<EOF>>	{ yyerror("unterminated quoted string"); }
+<xq,xd><<EOF>>	{ replication_yyerror("unterminated quoted string"); }
 
 
 <<EOF>>			{
@@ -231,7 +240,7 @@ addlitchar(unsigned char ychar)
 }
 
 void
-yyerror(const char *message)
+replication_yyerror(const char *message)
 {
 	ereport(ERROR,
 			(errcode(ERRCODE_SYNTAX_ERROR),

src/backend/replication/syncrep_gram.y

@@ -112,5 +112,3 @@ create_syncrep_config(const char *num_sync, List *members, uint8 syncrep_method)
 
 	return config;
 }
-
-#include "syncrep_scanner.c"

src/backend/replication/syncrep_scanner.l

@@ -1,4 +1,4 @@
-%{
+%top{
 /*-------------------------------------------------------------------------
  *
  * syncrep_scanner.l
@@ -17,6 +17,15 @@
 
 #include "lib/stringinfo.h"
 
+/*
+ * NB: include syncrep_gram.h only AFTER including syncrep.h, because syncrep.h
+ * includes node definitions needed for YYSTYPE.
+ */
+#include "replication/syncrep.h"
+#include "syncrep_gram.h"
+}
+
+%{
 /* Avoid exit() on fatal scanner errors (a bit ugly -- see yy_fatal_error) */
 #undef fprintf
 #define fprintf(file, fmt, msg)  fprintf_to_ereport(fmt, msg)
@@ -82,28 +91,28 @@ xdinside		[^"]+
 				appendStringInfoString(&xdbuf, yytext);
 		}
 <xd>{xdstop} {
-				yylval.str = xdbuf.data;
+				syncrep_yylval.str = xdbuf.data;
 				xdbuf.data = NULL;
 				BEGIN(INITIAL);
 				return NAME;
 		}
 <xd><<EOF>> {
-				yyerror("unterminated quoted identifier");
+				syncrep_yyerror("unterminated quoted identifier");
 				return JUNK;
 		}
 
 {identifier} {
-				yylval.str = pstrdup(yytext);
+				syncrep_yylval.str = pstrdup(yytext);
 				return NAME;
 		}
 
 {digit}+	{
-				yylval.str = pstrdup(yytext);
+				syncrep_yylval.str = pstrdup(yytext);
 				return NUM;
 		}
 
 "*"		{
-				yylval.str = "*";
+				syncrep_yylval.str = "*";
 				return NAME;
 		}
 

src/backend/utils/adt/.gitignore

@@ -1,2 +1,3 @@
+/jsonpath_gram.h
 /jsonpath_gram.c
 /jsonpath_scan.c

src/backend/utils/adt/Makefile

@@ -57,6 +57,7 @@ OBJS = \
 	jsonpath.o \
 	jsonpath_exec.o \
 	jsonpath_gram.o \
+	jsonpath_scan.o \
 	like.o \
 	like_support.o \
 	lockfuncs.o \
@@ -119,11 +120,17 @@ OBJS = \
 	xid8funcs.o \
 	xml.o
 
+# See notes in src/backend/parser/Makefile about the following two rules
+jsonpath_gram.h: jsonpath_gram.c
+	touch $@
+
+jsonpath_gram.c: BISONFLAGS += -d
+
 jsonpath_scan.c: FLEXFLAGS = -CF -p -p
 jsonpath_scan.c: FLEX_NO_BACKUP=yes
 
-# jsonpath_scan is compiled as part of jsonpath_gram
-jsonpath_gram.o: jsonpath_scan.c
+# Force these dependencies to be known even without dependency info built:
+jsonpath_gram.o jsonpath_scan.o: jsonpath_gram.h
 
 # jsonpath_gram.c and jsonpath_scan.c are in the distribution tarball,
 # so they are not cleaned here.

src/backend/utils/adt/jsonpath_gram.y

@@ -18,26 +18,11 @@
 
 #include "catalog/pg_collation.h"
 #include "fmgr.h"
+#include "jsonpath_internal.h"
 #include "miscadmin.h"
 #include "nodes/pg_list.h"
 #include "regex/regex.h"
 #include "utils/builtins.h"
-#include "utils/jsonpath.h"
-
-/* struct JsonPathString is shared between scan and gram */
-typedef struct JsonPathString
-{
-	char	   *val;
-	int			len;
-	int			total;
-}			JsonPathString;
-
-union YYSTYPE;
-
-/* flex 2.5.4 doesn't bother with a decl for this */
-int	jsonpath_yylex(union YYSTYPE *yylval_param);
-int	jsonpath_yyparse(JsonPathParseResult **result);
-void jsonpath_yyerror(JsonPathParseResult **result, const char *message);
 
 static JsonPathParseItem *makeItemType(JsonPathItemType type);
 static JsonPathParseItem *makeItemString(JsonPathString *s);
@@ -593,13 +578,3 @@ jspConvertRegexFlags(uint32 xflags)
 
 	return cflags;
 }
-
-/*
- * jsonpath_scan.l is compiled as part of jsonpath_gram.y.  Currently, this is
- * unavoidable because jsonpath_gram does not create a .h file to export its
- * token symbols.  If these files ever grow large enough to be worth compiling
- * separately, that could be fixed; but for now it seems like useless
- * complication.
- */
-
-#include "jsonpath_scan.c"

src/backend/utils/adt/jsonpath_internal.h

@@ -0,0 +1,32 @@
+/*-------------------------------------------------------------------------
+ *
+ * jsonpath_internal.h
+ *     Private definitions for jsonpath scanner & parser
+ *
+ * Portions Copyright (c) 1996-2022, PostgreSQL Global Development Group
+ * Portions Copyright (c) 1994, Regents of the University of California
+ *
+ * src/backend/utils/adt/jsonpath_internal.h
+ *
+ *-------------------------------------------------------------------------
+ */
+
+#ifndef JSONPATH_INTERNAL_H
+#define JSONPATH_INTERNAL_H
+
+/* struct JsonPathString is shared between scan and gram */
+typedef struct JsonPathString
+{
+	char	   *val;
+	int			len;
+	int			total;
+}			JsonPathString;
+
+#include "utils/jsonpath.h"
+#include "jsonpath_gram.h"
+
+extern int     jsonpath_yylex(YYSTYPE *yylval_param);
+extern int     jsonpath_yyparse(JsonPathParseResult **result);
+extern void jsonpath_yyerror(JsonPathParseResult **result, const char *message);
+
+#endif							/* JSONPATH_INTERNAL_H */

src/backend/utils/adt/jsonpath_scan.l

@@ -1,4 +1,4 @@
-%{
+%top{
 /*-------------------------------------------------------------------------
  *
  * jsonpath_scan.l
@@ -17,9 +17,18 @@
 
 #include "postgres.h"
 
+/*
+ * NB: include jsonpath_gram.h only AFTER including jsonpath_internal.h,
+ * because jsonpath_internal.h contains the declaration for JsonPathString.
+ */
+#include "jsonpath_internal.h"
+#include "jsonpath_gram.h"
+
 #include "mb/pg_wchar.h"
 #include "nodes/pg_list.h"
+}
 
+%{
 static JsonPathString scanstring;
 
 /* Handles to the buffer that the lexer uses internally */
@@ -142,9 +151,9 @@ hex_fail	\\x{hex_dig}{0,1}
 
 <xnq,xq,xvq>{hex_char}		{ parseHexChar(yytext); }
 
-<xnq,xq,xvq>{unicode}*{unicodefail}	{ yyerror(NULL, "invalid unicode sequence"); }
+<xnq,xq,xvq>{unicode}*{unicodefail}	{ jsonpath_yyerror(NULL, "invalid unicode sequence"); }
 
-<xnq,xq,xvq>{hex_fail}		{ yyerror(NULL, "invalid hex character sequence"); }
+<xnq,xq,xvq>{hex_fail}		{ jsonpath_yyerror(NULL, "invalid hex character sequence"); }
 
 <xnq,xq,xvq>{unicode}+\\	{
 								/* throw back the \\, and treat as unicode */
@@ -154,9 +163,9 @@ hex_fail	\\x{hex_dig}{0,1}
 
 <xnq,xq,xvq>\\.				{ addchar(false, yytext[1]); }
 
-<xnq,xq,xvq>\\				{ yyerror(NULL, "unexpected end after backslash"); }
+<xnq,xq,xvq>\\				{ jsonpath_yyerror(NULL, "unexpected end after backslash"); }
 
-<xq,xvq><<EOF>>				{ yyerror(NULL, "unexpected end of quoted string"); }
+<xq,xvq><<EOF>>				{ jsonpath_yyerror(NULL, "unexpected end of quoted string"); }
 
 <xq>\"							{
 									yylval->str = scanstring;
@@ -178,7 +187,7 @@ hex_fail	\\x{hex_dig}{0,1}
 
 <xc>\*							{ }
 
-<xc><<EOF>>						{ yyerror(NULL, "unexpected end of comment"); }
+<xc><<EOF>>						{ jsonpath_yyerror(NULL, "unexpected end of comment"); }
 
 \&\&							{ return AND_P; }
 
@@ -244,10 +253,10 @@ hex_fail	\\x{hex_dig}{0,1}
 									return INT_P;
 								}
 
-{realfail}						{ yyerror(NULL, "invalid numeric literal"); }
-{integer_junk}					{ yyerror(NULL, "trailing junk after numeric literal"); }
-{decimal_junk}					{ yyerror(NULL, "trailing junk after numeric literal"); }
-{real_junk}						{ yyerror(NULL, "trailing junk after numeric literal"); }
+{realfail}						{ jsonpath_yyerror(NULL, "invalid numeric literal"); }
+{integer_junk}					{ jsonpath_yyerror(NULL, "trailing junk after numeric literal"); }
+{decimal_junk}					{ jsonpath_yyerror(NULL, "trailing junk after numeric literal"); }
+{real_junk}						{ jsonpath_yyerror(NULL, "trailing junk after numeric literal"); }
 
 \"								{
 									addchar(true, '\0');

src/backend/utils/misc/Makefile

@@ -16,6 +16,7 @@ override CPPFLAGS := -I. -I$(srcdir) $(CPPFLAGS)
 
 OBJS = \
 	guc.o \
+	guc-file.o \
 	help_config.o \
 	pg_config.o \
 	pg_controldata.o \
@@ -37,10 +38,6 @@ endif
 
 include $(top_srcdir)/src/backend/common.mk
 
-# guc-file is compiled as part of guc
-guc.o: guc-file.c
-
 # Note: guc-file.c is not deleted by 'make clean',
 # since we want to ship it in distribution tarballs.
 clean:
-	@rm -f lex.yy.c

src/backend/utils/misc/guc-file.l

@@ -1,4 +1,4 @@
-/* -*-pgsql-c-*- */
+%top{
 /*
  * Scanner for the configuration file
  *
@@ -7,8 +7,6 @@
  * src/backend/utils/misc/guc-file.l
  */
 
-%{
-
 #include "postgres.h"
 
 #include <ctype.h>
@@ -19,8 +17,12 @@
 #include "mb/pg_wchar.h"
 #include "miscadmin.h"
 #include "storage/fd.h"
+#include <sys/stat.h>
+#include "utils/memutils.h"
+}
 
 
+%{
 /*
  * flex emits a yy_fatal_error() function that it calls in response to
  * critical errors like malloc failure, file I/O errors, and detection of

src/backend/utils/misc/guc.c

@@ -13333,5 +13333,3 @@ check_default_with_oids(bool *newval, void **extra, GucSource source)
 
 	return true;
 }
-
-#include "guc-file.c"

src/bin/pgbench/.gitignore

@@ -1,3 +1,4 @@
+/exprparse.h
 /exprparse.c
 /exprscan.c
 /pgbench

src/bin/pgbench/Makefile

@@ -10,6 +10,7 @@ include $(top_builddir)/src/Makefile.global
 OBJS = \
 	$(WIN32RES) \
 	exprparse.o \
+	exprscan.o \
 	pgbench.o
 
 override CPPFLAGS := -I. -I$(srcdir) -I$(libpq_srcdir) $(CPPFLAGS)
@@ -26,8 +27,14 @@ all: pgbench
 pgbench: $(OBJS) | submake-libpq submake-libpgport submake-libpgfeutils
 	$(CC) $(CFLAGS) $^ $(LDFLAGS) $(LDFLAGS_EX) $(LIBS) -o $@$(X)
 
-# exprscan is compiled as part of exprparse
-exprparse.o: exprscan.c
+# See notes in src/backend/parser/Makefile about the following two rules
+exprparse.h: exprparse.c
+	touch $@
+
+exprparse.c: BISONFLAGS += -d
+
+# Force these dependencies to be known even without dependency info built:
+exprparse.o exprscan.o: exprparse.h
 
 distprep: exprparse.c exprscan.c
 
@@ -45,7 +52,7 @@ clean distclean:
 	rm -rf tmp_check
 
 maintainer-clean: distclean
-	rm -f exprparse.c exprscan.c
+	rm -f exprparse.h exprparse.c exprscan.c
 
 check:
 	$(prove_check)

src/bin/pgbench/exprparse.y

@@ -526,18 +526,3 @@ make_case(yyscan_t yyscanner, PgBenchExprList *when_then_list, PgBenchExpr *else
 					 find_func(yyscanner, "!case_end"),
 					 make_elist(else_part, when_then_list));
 }
-
-/*
- * exprscan.l is compiled as part of exprparse.y.  Currently, this is
- * unavoidable because exprparse does not create a .h file to export
- * its token symbols.  If these files ever grow large enough to be
- * worth compiling separately, that could be fixed; but for now it
- * seems like useless complication.
- */
-
-/* First, get rid of "#define yyscan_t" from pgbench.h */
-#undef yyscan_t
-/* ... and the yylval macro, which flex will have its own definition for */
-#undef yylval
-
-#include "exprscan.c"

src/bin/pgbench/exprscan.l

@@ -1,4 +1,4 @@
-%{
+%top{
 /*-------------------------------------------------------------------------
  *
  * exprscan.l
@@ -22,9 +22,19 @@
  *
  *-------------------------------------------------------------------------
  */
+#include "postgres_fe.h"
 
+/*
+ * NB: include exprparse.h only AFTER including pgbench.h, because pgbench.h
+ * contains definitions needed for YYSTYPE. Likewise, pgbench.h must come after
+ * psqlscan_int.h for yyscan_t.
+ */
 #include "fe_utils/psqlscan_int.h"
+#include "pgbench.h"
+#include "exprparse.h"
+}
 
+%{
 /* context information for reporting errors in expressions */
 static const char *expr_source = NULL;
 static int	expr_lineno = 0;

src/test/isolation/.gitignore

@@ -3,6 +3,7 @@
 /pg_isolation_regress
 
 # Local generated source files
+/specparse.h
 /specparse.c
 /specscanner.c
 

src/test/isolation/Makefile

@@ -15,7 +15,8 @@ override CPPFLAGS := -I. -I$(srcdir) -I$(libpq_srcdir) \
 OBJS = \
 	$(WIN32RES) \
 	isolationtester.o \
-	specparse.o
+	specparse.o \
+	specscanner.o
 
 all: isolationtester$(X) pg_isolation_regress$(X)
 
@@ -44,8 +45,14 @@ isolationtester$(X): $(OBJS) | submake-libpq submake-libpgport
 
 distprep: specparse.c specscanner.c
 
-# specscanner is compiled as part of specparse
-specparse.o: specscanner.c
+# See notes in src/backend/parser/Makefile about the following two rules
+specparse.h: specparse.c
+	touch $@
+
+specparse.c: BISONFLAGS += -d
+
+# Force these dependencies to be known even without dependency info built:
+specparse.o specscanner.o: specparse.h
 
 # specparse.c and specscanner.c are in the distribution tarball,
 # so do not clean them here
@@ -55,7 +62,7 @@ clean distclean:
 	rm -rf $(pg_regress_clean_files)
 
 maintainer-clean: distclean
-	rm -f specparse.c specscanner.c
+	rm -f specparse.h specparse.c specscanner.c
 
 installcheck: all
 	$(pg_isolation_regress_installcheck) --schedule=$(srcdir)/isolation_schedule

src/test/isolation/specparse.y

@@ -276,5 +276,3 @@ blocker:
 		;
 
 %%
-
-#include "specscanner.c"

src/test/isolation/specscanner.l

@@ -1,4 +1,4 @@
-%{
+%top{
 /*-------------------------------------------------------------------------
  *
  * specscanner.l
@@ -9,7 +9,17 @@
  *
  *-------------------------------------------------------------------------
  */
+#include "postgres_fe.h"
 
+/*
+ * NB: include specparse.h only AFTER including isolationtester.h, because
+ * isolationtester.h includes node definitions needed for YYSTYPE.
+ */
+#include "isolationtester.h"
+#include "specparse.h"
+}
+
+%{
 static int	yyline = 1;			/* line number for error reporting */
 
 #define LITBUF_INIT	1024		/* initial size of litbuf */
@@ -75,7 +85,7 @@ teardown		{ return TEARDOWN; }
 
  /* Plain identifiers */
 {identifier}	{
-					yylval.str = pg_strdup(yytext);
+					spec_yylval.str = pg_strdup(yytext);
 					return(identifier);
 				}
 
@@ -87,13 +97,13 @@ teardown		{ return TEARDOWN; }
 <qident>\"\"	{ addlitchar(yytext[0]); }
 <qident>\"		{
 					litbuf[litbufpos] = '\0';
-					yylval.str = pg_strdup(litbuf);
+					spec_yylval.str = pg_strdup(litbuf);
 					BEGIN(INITIAL);
 					return(identifier);
 				}
 <qident>.		{ addlitchar(yytext[0]); }
-<qident>\n		{ yyerror("unexpected newline in quoted identifier"); }
-<qident><<EOF>>	{ yyerror("unterminated quoted identifier"); }
+<qident>\n		{ spec_yyerror("unexpected newline in quoted identifier"); }
+<qident><<EOF>>	{ spec_yyerror("unterminated quoted identifier"); }
 
  /* SQL blocks: { UPDATE ... } */
  /* We trim leading/trailing whitespace, otherwise they're unprocessed */
@@ -104,7 +114,7 @@ teardown		{ return TEARDOWN; }
 				}
 <sql>{space}*"}" {
 					litbuf[litbufpos] = '\0';
-					yylval.str = pg_strdup(litbuf);
+					spec_yylval.str = pg_strdup(litbuf);
 					BEGIN(INITIAL);
 					return(sqlblock);
 				}
@@ -116,12 +126,12 @@ teardown		{ return TEARDOWN; }
 					addlitchar(yytext[0]);
 				}
 <sql><<EOF>>	{
-					yyerror("unterminated sql block");
+					spec_yyerror("unterminated sql block");
 				}
 
  /* Numbers and punctuation */
 {digit}+		{
-					yylval.integer = atoi(yytext);
+					spec_yylval.integer = atoi(yytext);
 					return INTEGER;
 				}
 
@@ -150,7 +160,7 @@ addlitchar(char c)
 }
 
 void
-yyerror(const char *message)
+spec_yyerror(const char *message)
 {
 	fprintf(stderr, "%s at line %d\n", message, yyline);
 	exit(1);

src/tools/pginclude/headerscheck

@@ -117,10 +117,18 @@ do
 	# We can't make these Bison output files compilable standalone
 	# without using "%code require", which old Bison versions lack.
 	# parser/gram.h will be included by parser/gramparse.h anyway.
+	test "$f" = contrib/cube/cubeparse.h && continue
+	test "$f" = contrib/seg/segparse.h && continue
+	test "$f" = src/backend/bootstrap/bootparse.h && continue
 	test "$f" = src/include/parser/gram.h && continue
 	test "$f" = src/backend/parser/gram.h && continue
+	test "$f" = src/backend/replication/repl_gram.h && continue
+	test "$f" = src/backend/replication/syncrep_gram.h && continue
+	test "$f" = src/backend/utils/adt/jsonpath_gram.h && continue
+	test "$f" = src/bin/pgbench/exprparse.h && continue
 	test "$f" = src/pl/plpgsql/src/pl_gram.h && continue
 	test "$f" = src/interfaces/ecpg/preproc/preproc.h && continue
+	test "$f" = src/test/isolation/specparse.h && continue
 
 	# This produces a "no previous prototype" warning.
 	test "$f" = src/include/storage/checksum_impl.h && continue