Add hts_parse_decimal() and hts_parse_region() flags parameter [DRAFT]

Add the first flag, HTS_PARSE_THOUSANDS_SEP.

[IN PROGRESS]  Need to figure out whether hts_parse_region() is workable
with a strend argument and the possibility of colons in chromosome names...

Author: jmarshall

hts.c

@@ -1824,7 +1824,7 @@ static inline long long push_digit(long long i, char c)
     return 10 * i + digit;
 }
 
-long long hts_parse_decimal(const char *str, char **end)
+long long hts_parse_decimal(const char *str, char **strend, int flags)
 {
     long long n = 0;
     int decimals = 0, e = 0, lost = 0;
@@ -1837,7 +1837,7 @@ long long hts_parse_decimal(const char *str, char **end)
     if (*s == '+' || *s == '-') sign = *s++;
     while (*s)
         if (isdigit(*s)) n = push_digit(n, *s++);
-        else if (*s == ',') s++;
+        else if (*s == ',' && (flags & HTS_PARSE_THOUSANDS_SEP)) s++;
         else break;
 
     if (*s == '.') {
@@ -1860,7 +1860,7 @@ long long hts_parse_decimal(const char *str, char **end)
         fprintf(stderr, "[W::%s] discarding fractional part of %.*s\n",
                 __func__, (int)(s - str), str);
 
-    if (end) *end = (char *) s;
+    if (strend) *strend = (char *) s;
     else if (*s && hts_verbose >= 2)
         fprintf(stderr, "[W::%s] ignoring unknown characters after %.*s[%s]\n",
                 __func__, (int)(s - str), str, s);
@@ -1869,6 +1869,12 @@ long long hts_parse_decimal(const char *str, char **end)
 }
 
 const char *hts_parse_reg(const char *s, int *beg, int *end)
+{
+    return hts_parse_region(s, NULL, beg, end, HTS_PARSE_THOUSANDS_SEP);
+}
+
+const char *
+hts_parse_region(const char *s, char **strend, int *beg, int *end, int flags)
 {
     char *hyphen;
     const char *colon = strrchr(s, ':');
@@ -1877,11 +1883,12 @@ const char *hts_parse_reg(const char *s, int *beg, int *end)
         return s + strlen(s);
     }
 
-    *beg = hts_parse_decimal(colon+1, &hyphen) - 1;
+    *beg = hts_parse_decimal(colon+1, &hyphen, flags) - 1;
     if (*beg < 0) *beg = 0;
 
+    // FIXME \0 vs. return NULL
     if (*hyphen == '\0') *end = INT_MAX;
-    else if (*hyphen == '-') *end = hts_parse_decimal(hyphen+1, NULL);
+    else if (*hyphen == '-') *end = hts_parse_decimal(hyphen+1, strend, flags);
     else return NULL;
 
     if (*beg >= *end) return NULL;

htslib/hts.h

@@ -477,27 +477,40 @@ hts_idx_t *hts_idx_load2(const char *fn, const char *fnidx);
     int hts_idx_get_stat(const hts_idx_t* idx, int tid, uint64_t* mapped, uint64_t* unmapped);
     uint64_t hts_idx_get_n_no_coor(const hts_idx_t* idx);
 
+
+#define HTS_PARSE_THOUSANDS_SEP 1  ///< Ignore ',' separators within numbers
+
 /// Parse a numeric string
-/** The number may be expressed in scientific notation, and may contain commas
-    in the integer part (before any decimal point or E notation).
-    @param str  String to be parsed
-    @param end  If non-NULL, set on return to point to the first character
-                in @a str after those forming the parsed number
+/** The number may be expressed in scientific notation, and optionally may
+    contain commas in the integer part (before any decimal point or E notation).
+    @param str     String to be parsed
+    @param strend  If non-NULL, set on return to point to the first character
+                   in @a str after those forming the parsed number
+    @param flags   Or'ed-together combination of HTS_PARSE_* flags
     @return  Converted value of the parsed number.
 
-    When @a end is NULL, a warning will be printed (if hts_verbose is 2
+    When @a strend is NULL, a warning will be printed (if hts_verbose is 2
     or more) if there are any trailing characters after the number.
 */
-long long hts_parse_decimal(const char *str, char **end);
+long long hts_parse_decimal(const char *str, char **strend, int flags);
+
+/// Equivalent to hts_parse_region(str, NULL, beg, end, HTS_PARSE_THOUSANDS_SEP)
+const char *hts_parse_reg(const char *str, int *beg, int *end);
 
 /// Parse a "CHR:START-END"-style region string
-/** @param str  String to be parsed
-    @param beg  Set on return to the 0-based start of the region
-    @param end  Set on return to the 1-based end of the region
-    @return  Pointer to the colon or '\0' after the reference sequence name,
-             or NULL if @a str could not be parsed.
+/** @param str     String to be parsed
+    @param strend  If non-NULL, set on return to point to the first character
+                   in @a str after those forming the parsed region
+    @param beg     Set on return to the 0-based start of the region
+    @param end     Set on return to the 1-based end of the region
+    @param flags   Or'ed-together combination of HTS_PARSE_* flags
+    @return  Pointer to the colon or terminating character after the reference
+             sequence name, or NULL if @a str could not be parsed.
+
+    When @a strend is NULL, a warning will be printed (if hts_verbose is 2
+    or more) if there are any trailing characters after the region string.
 */
-const char *hts_parse_reg(const char *str, int *beg, int *end);
+const char *hts_parse_region(const char *str, char **strend, int *beg, int *end, int flags);
 
     hts_itr_t *hts_itr_query(const hts_idx_t *idx, int tid, int beg, int end, hts_readrec_func *readrec);
     void hts_itr_destroy(hts_itr_t *iter);

regidx.c

@@ -297,11 +297,11 @@ int regidx_parse_bed(const char *line, char **chr_beg, char **chr_end, reg_t *re
     *chr_end = se-1;
 
     ss = se+1;
-    reg->start = hts_parse_decimal(ss, &se);
+    reg->start = hts_parse_decimal(ss, &se, 0);
     if ( ss==se ) { fprintf(stderr,"Could not parse bed line: %s\n", line); return -2; }
 
     ss = se+1;
-    reg->end = hts_parse_decimal(ss, &se) - 1;
+    reg->end = hts_parse_decimal(ss, &se, 0) - 1;
     if ( ss==se ) { fprintf(stderr,"Could not parse bed line: %s\n", line); return -2; }
 
     return 0;
@@ -322,15 +322,15 @@ int regidx_parse_tab(const char *line, char **chr_beg, char **chr_end, reg_t *re
     *chr_end = se-1;
 
     ss = se+1;
-    reg->start = hts_parse_decimal(ss, &se) - 1;
+    reg->start = hts_parse_decimal(ss, &se, 0) - 1;
     if ( ss==se ) { fprintf(stderr,"Could not parse bed line: %s\n", line); return -2; }
 
     if ( !se[0] || !se[1] )
         reg->end = reg->start;
     else
     {
         ss = se+1;
-        reg->end = hts_parse_decimal(ss, &se);
+        reg->end = hts_parse_decimal(ss, &se, 0);
         if ( ss==se ) reg->end = reg->start;
         else reg->end--;
     }

synced_bcf_reader.c

@@ -887,7 +887,7 @@ static bcf_sr_regions_t *_regions_init_string(const char *str)
         if ( *ep==':' )
         {
             sp = ep+1;
-            from = hts_parse_decimal(sp,(char**)&ep);
+            from = hts_parse_decimal(sp,(char**)&ep,0);
             if ( sp==ep )
             {
                 fprintf(stderr,"[%s:%d %s] Could not parse the region(s): %s\n", __FILE__,__LINE__,__FUNCTION__,str);
@@ -906,7 +906,7 @@ static bcf_sr_regions_t *_regions_init_string(const char *str)
             }
             ep++;
             sp = ep;
-            to = hts_parse_decimal(sp,(char**)&ep);
+            to = hts_parse_decimal(sp,(char**)&ep,0);
             if ( *ep && *ep!=',' )
             {
                 fprintf(stderr,"[%s:%d %s] Could not parse the region(s): %s\n", __FILE__,__LINE__,__FUNCTION__,str);
@@ -953,15 +953,15 @@ static int _regions_parse_line(char *line, int ichr,int ifrom,int ito, char **ch
     if ( i<=k ) return -1;
     if ( k==l )
     {
-        *from = *to = hts_parse_decimal(ss, &tmp);
+        *from = *to = hts_parse_decimal(ss, &tmp, 0);
         if ( tmp==ss ) return -1;
     }
     else
     {
         if ( k==ifrom )
-            *from = hts_parse_decimal(ss, &tmp);
+            *from = hts_parse_decimal(ss, &tmp, 0);
         else
-            *to = hts_parse_decimal(ss, &tmp);
+            *to = hts_parse_decimal(ss, &tmp, 0);
         if ( ss==tmp ) return -1;
 
         for (i=k; i<l && *se; i++)
@@ -971,9 +971,9 @@ static int _regions_parse_line(char *line, int ichr,int ifrom,int ito, char **ch
         }
         if ( i<l ) return -1;
         if ( k==ifrom )
-            *to = hts_parse_decimal(ss, &tmp);
+            *to = hts_parse_decimal(ss, &tmp, 0);
         else
-            *from = hts_parse_decimal(ss, &tmp);
+            *from = hts_parse_decimal(ss, &tmp, 0);
         if ( ss==tmp ) return -1;
     }