Improve error reporting for problems in text search configuration files

by installing an error context subroutine that will provide the file name
and line number for all errors detected while reading a config file.
Some of the reader routines were already doing that in an ad-hoc way for
errors detected directly in the reader, but it didn't help for problems
detected in subroutines, such as encoding violations.

Back-patch to 8.3 because 8.3 is where people will be trying to debug
configuration files.

1 files changed, 103 insertions, 1 deletions

diff --git a/src/backend/tsearch/ts_locale.c b/src/backend/tsearch/ts_locale.c
index dacf7325c6e..73d8ed51de7 100644
--- a/src/backend/tsearch/ts_locale.c
+++ b/src/backend/tsearch/ts_locale.c
@@ -7,15 +7,18 @@
  *
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/backend/tsearch/ts_locale.c,v 1.7 2008/01/01 19:45:52 momjian Exp $
+ *	  $PostgreSQL: pgsql/src/backend/tsearch/ts_locale.c,v 1.7.2.1 2008/06/18 20:55:49 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
 #include "postgres.h"
 
+#include "storage/fd.h"
 #include "tsearch/ts_locale.h"
 #include "tsearch/ts_public.h"
 
+static void tsearch_readline_callback(void *arg);
+
 
 #ifdef TS_USE_WIDE
 
@@ -194,9 +197,108 @@ t_isprint(const char *ptr)
 
 
 /*
+ * Set up to read a file using tsearch_readline().  This facility is
+ * better than just reading the file directly because it provides error
+ * context pointing to the specific line where a problem is detected.
+ *
+ * Expected usage is:
+ *
+ *		tsearch_readline_state trst;
+ *
+ *		if (!tsearch_readline_begin(&trst, filename))
+ *			ereport(ERROR,
+ *					(errcode(ERRCODE_CONFIG_FILE_ERROR),
+ *					 errmsg("could not open stop-word file \"%s\": %m",
+ *							filename)));
+ *		while ((line = tsearch_readline(&trst)) != NULL)
+ *			process line;
+ *		tsearch_readline_end(&trst);
+ *
+ * Note that the caller supplies the ereport() for file open failure;
+ * this is so that a custom message can be provided.  The filename string
+ * passed to tsearch_readline_begin() must remain valid through
+ * tsearch_readline_end().
+ */
+bool
+tsearch_readline_begin(tsearch_readline_state *stp,
+					   const char *filename)
+{
+	if ((stp->fp = AllocateFile(filename, "r")) == NULL)
+		return false;
+	stp->filename = filename;
+	stp->lineno = 0;
+	stp->curline = NULL;
+	/* Setup error traceback support for ereport() */
+	stp->cb.callback = tsearch_readline_callback;
+	stp->cb.arg = (void *) stp;
+	stp->cb.previous = error_context_stack;
+	error_context_stack = &stp->cb;
+	return true;
+}
+
+/*
+ * Read the next line from a tsearch data file (expected to be in UTF-8), and
+ * convert it to database encoding if needed. The returned string is palloc'd.
+ * NULL return means EOF.
+ */
+char *
+tsearch_readline(tsearch_readline_state *stp)
+{
+	char	   *result;
+
+	stp->lineno++;
+	stp->curline = NULL;
+	result = t_readline(stp->fp);
+	stp->curline = result;
+	return result;
+}
+
+/*
+ * Close down after reading a file with tsearch_readline()
+ */
+void
+tsearch_readline_end(tsearch_readline_state *stp)
+{
+	FreeFile(stp->fp);
+	/* Pop the error context stack */
+	error_context_stack = stp->cb.previous;
+}
+
+/*
+ * Error context callback for errors occurring while reading a tsearch
+ * configuration file.
+ */
+static void
+tsearch_readline_callback(void *arg)
+{
+	tsearch_readline_state *stp = (tsearch_readline_state *) arg;
+
+	/*
+	 * We can't include the text of the config line for errors that occur
+	 * during t_readline() itself.  This is only partly a consequence of
+	 * our arms-length use of that routine: the major cause of such
+	 * errors is encoding violations, and we daren't try to print error
+	 * messages containing badly-encoded data.
+	 */
+	if (stp->curline)
+		errcontext("line %d of configuration file \"%s\": \"%s\"",
+				   stp->lineno,
+				   stp->filename,
+				   stp->curline);
+	else
+		errcontext("line %d of configuration file \"%s\"",
+				   stp->lineno,
+				   stp->filename);
+}
+
+
+/*
  * Read the next line from a tsearch data file (expected to be in UTF-8), and
  * convert it to database encoding if needed. The returned string is palloc'd.
  * NULL return means EOF.
+ *
+ * Note: direct use of this function is now deprecated.  Go through
+ * tsearch_readline() to provide better error reporting.
  */
 char *
 t_readline(FILE *fp)