misc refactorings (#142)

* move zsv_strencode.c to standalone file
change utils/dirs_from_json to accommodate backslash path separators
modify zsv_foreach_dirent to stop sooner if max_depth provided and to abort if subdir handler returns err

app/utils/dirs.c

@@ -19,6 +19,7 @@
 #include <zsv/utils/string.h>
 #include <unistd.h> // unlink
 #include <sys/stat.h>
+#include <yajl_helper/yajl_helper.h>
 
 #if defined(_WIN32)
 #include <windows.h>
@@ -245,7 +246,7 @@ int zsv_foreach_dirent_aux(const char *dir_path,
  if(!dir_path)
  return 1;
 
- if(max_depth > 0 && depth > max_depth)
+ if(max_depth > 0 && depth >= max_depth)
  return 0;
 
  DIR *dr;
@@ -269,11 +270,11 @@ int zsv_foreach_dirent_aux(const char *dir_path,
  char is_dir = h.stat.st_mode & S_IFDIR ? 1 : 0;
  h.is_dir = is_dir;
  if(handler)
- err = handler(&h, depth + 1);
+ handler(&h, depth + 1);
 
  if(is_dir && !h.no_recurse)
  // recurse!
- zsv_foreach_dirent_aux(tmp, depth + 1, max_depth, handler, ctx, verbose);
+ err = zsv_foreach_dirent_aux(tmp, depth + 1, max_depth, handler, ctx, verbose);
  free(tmp);
  }
  }
@@ -312,6 +313,9 @@ int zsv_remove_dir_recursive(const unsigned char *path) {
  return err;
 }
 
+#ifndef ZSV_NO_JQ
 #include "dirs_to_json.c"
 
 #include "dirs_from_json.c"
+
+#endif

app/utils/dirs_from_json.c

@@ -41,6 +41,10 @@ static int zsv_dir_from_json_map_key(struct yajl_helper_parse_state *st,
  asprintf(&fn, "%s%c%.*s", ctx->filepath_prefix, FILESLASH, (int)len, s);
  else
  asprintf(&fn, "%.*s", (int)len, s);
+
+ // if we have any backslashes, replace with fwd slash
+ if(fn)
+ for(int i = 0, j = strlen(fn); i < j; i++) if(fn[i] == '\\') fn[i] = '/';
  if(!fn) {
  errno = ENOMEM;
  perror(NULL);

app/utils/string.c

@@ -16,6 +16,10 @@
 #include <zsv/utils/utf8.h>
 #include <zsv/utils/string.h>
 
+#ifdef ZSV_UTILS_STRING_STANDALONE
+#include "../../src/zsv_strencode.c"
+#endif
+
 #ifndef NO_UTF8PROC
 #include <utf8proc.h>
 
@@ -326,8 +330,10 @@ size_t zsv_strunescape_backslash(unsigned char *s, size_t len) {
  return j;
 }
 
+#ifndef ZSV_STRING_LIB_ONLY
 struct zsv_cell zsv_get_cell_trimmed(zsv_parser parser, size_t ix) {
  struct zsv_cell c = zsv_get_cell(parser, ix);
  c.str = (unsigned char *)zsv_strtrim(c.str, &c.len);
  return c;
 }
+#endif

src/zsv.c

@@ -11,7 +11,6 @@
 #endif
 
 #include "zsv.h"
-#include <zsv/utils/utf8.h>
 #include <zsv/utils/compiler.h>
 #ifdef ZSV_EXTRAS
 #include <zsv/utils/arg.h>
@@ -30,45 +29,7 @@ const char *zsv_lib_version(void) {
  return ZSV_VERSION;
 }
 
-/**
- * Ensure valid UTF8 encoding by, if needed, replacing malformed bytes
- */
-ZSV_EXPORT
-size_t zsv_strencode(unsigned char *s, size_t n, unsigned char replace,
- int (*malformed_handler)(void *, const unsigned char *s, size_t n, size_t offset), void *handler_ctx) {
- size_t new_len = 0;
- int clen;
- for(size_t i2 = 0; i2 < n; i2 += (size_t)clen) {
- clen = ZSV_UTF8_CHARLEN(s[i2]);
- if(LIKELY(clen == 1))
- s[new_len++] = s[i2];
- else if(UNLIKELY(clen < 0) || UNLIKELY(i2 + clen >= n)) {
- if(malformed_handler)
- malformed_handler(handler_ctx, s, n, new_len);
- if(replace)
- s[new_len++] = replace;
- clen = 1;
- } else { /* might be valid multi-byte utf8; check */
- unsigned char valid_n;
- for(valid_n = 1; valid_n < clen; valid_n++)
- if(!ZSV_UTF8_SUBSEQUENT_CHAR_OK(s[i2 + valid_n]))
- break;
- if(valid_n == clen) { /* valid_n utf8; copy it */
- memmove(s + new_len, s + i2, clen);
- new_len += clen;
- } else { /* invalid; valid_n smaller than expected */
- if(malformed_handler)
- malformed_handler(handler_ctx, s, n, new_len);
- if(replace) {
- memset(s + new_len, replace, valid_n);
- new_len += valid_n;
- }
- clen = valid_n;
- }
- }
- }
- return new_len; // new length
-}
+#include "zsv_strencode.c"
 
 /**
  * When we parse a chunk, if it was not the first parse call, we might have a partial

src/zsv_strencode.c

@@ -0,0 +1,49 @@
+/*
+ * zsv_strencode(): standalone file to allow zsv utilities that use this
+ * to be used on a standalone basis without the zsv parser
+ *
+ * This file is part of zsv/lib, distributed under the license defined at
+ * https://opensource.org/licenses/MIT
+ */
+
+#include <zsv/utils/utf8.h>
+#include <zsv/utils/compiler.h>
+/**
+ * Ensure valid UTF8 encoding by, if needed, replacing malformed bytes
+ */
+ZSV_EXPORT
+size_t zsv_strencode(unsigned char *s, size_t n, unsigned char replace,
+ int (*malformed_handler)(void *, const unsigned char *s, size_t n, size_t offset), void *handler_ctx) {
+ size_t new_len = 0;
+ int clen;
+ for(size_t i2 = 0; i2 < n; i2 += (size_t)clen) {
+ clen = ZSV_UTF8_CHARLEN(s[i2]);
+ if(LIKELY(clen == 1))
+ s[new_len++] = s[i2];
+ else if(UNLIKELY(clen < 0) || UNLIKELY(i2 + clen >= n)) {
+ if(malformed_handler)
+ malformed_handler(handler_ctx, s, n, new_len);
+ if(replace)
+ s[new_len++] = replace;
+ clen = 1;
+ } else { /* might be valid multi-byte utf8; check */
+ unsigned char valid_n;
+ for(valid_n = 1; valid_n < clen; valid_n++)
+ if(!ZSV_UTF8_SUBSEQUENT_CHAR_OK(s[i2 + valid_n]))
+ break;
+ if(valid_n == clen) { /* valid_n utf8; copy it */
+ memmove(s + new_len, s + i2, clen);
+ new_len += clen;
+ } else { /* invalid; valid_n smaller than expected */
+ if(malformed_handler)
+ malformed_handler(handler_ctx, s, n, new_len);
+ if(replace) {
+ memset(s + new_len, replace, valid_n);
+ new_len += valid_n;
+ }
+ clen = valid_n;
+ }
+ }
+ }
+ return new_len; // new length
+}