* Improved directory handling.

* Improved error handling and reporting.
* Corrected command line usage synopsis.
* Corrected old spelling (imdupe) to new (imgdupe) in README.md.
* Minor edits in README.md.

diff --git a/README.md b/README.md
index 687107678c2c8d1a530a22597524693f4b5e5f2d..4b5315d471e033adb8853bdb8657c3c4b2678dbf 100644 (file)
--- a/README.md
+++ b/README.md
@@ -1,23 +1,24 @@
-# Imdupe
+# Imgdupe
 
-Imdupe is a command line tool to find potential duplicate images in a
+Imgdupe is a command line tool to find potential duplicate images in a
 directory tree by comparing perceptual hash values of image files.  It
 writes a shell script to stdout which can in turn be used to further
 process the scan results.
 
-The perceptual hash for an image is calculated by:
+The perceptual hash for an image is calculated by applying the following
+transformations in the given order:
 
-  * flattening the image (merging layers/frames)
-  * transforming the color space to gray scale
-  * resizing to 16 by 16 pixel using box blur
-  * normalizing contrast and brightness
-  * conversion to monochromatic image
-  * interpreting the pixel values as 256 bit hash value
+  * flatten image (merge layers/frames)
+  * transform color space to gray scale
+  * resize to 16 by 16 pixel using box blur
+  * normalize
+  * convert to hard black and white
+  * interpret the resulting pixel values as 256 bit hash value
 
 
 ## Usage
 ```
-imgdupe [OPTIONS] DIR ...
+imgdupe [OPTIONS] [DIR ...]
 
 OPTIONS:
   -d n      max directory recursion depth; default: 0 (unlimited)
@@ -33,8 +34,11 @@ OPTIONS:
 ```
 **Notes**
 
+  * If no directory is specified on the command line, `imgdupe` scans
+    the current working directory.
+
   * `-d` limits the recursion depth when scanning a directory tree; a
-    value of 1 causes `imdupe` to only scan files in the directories
+    value of 1 causes `imgdupe` to only scan files in the directories
     specified on the command line and not descend into any sub-directories,
     0 (the default) means unlimited depth.
 
@@ -62,7 +66,7 @@ OPTIONS:
 
   * `-I` allows to include a custom lead-in in the output, which is in
     particular useful to replace the default dummy versions of the `VIEW()`
-    and `END()` functions which are used by `imdupe` to structure its
+    and `END()` functions which are used by `imgdupe` to structure its
     output.
 
   * `-T` allows to set the number of hashing threads to use; for optimal
@@ -73,10 +77,10 @@ OPTIONS:
 ## Build
 
 The [GraphicsMagick](http://www.graphicsmagick.org/) library must be
-installed as a prerequisite prior to building `imdupe`, as it is required
+installed as a prerequisite prior to building `imgdupe`, as it is required
 to manipulate images during perceptual hash calculation.
 
-Run `make` in the project directory to build the `imdupe` executable.
+Run GNU `make` in the project directory to build the `imgdupe` executable.
 
 Though it was only tested on GNU/Linux, it should work with few (if any)
 modifications on other Unix-like systems.
@@ -90,7 +94,7 @@ are only portable between systems of equal endianness.
 
 ## License
 
-Imdupe is distributed under the Modified ("3-clause") BSD License.
+Imgdupe is distributed under the Modified ("3-clause") BSD License.
 See `LICENSE` file for more information.
 
 ----------------------------------------------------------------------

diff --git a/main.c b/main.c
index 7d5d1b3abaef646553a192267c16be4953f66578..426a253a86c67bc139268cdb90cef6bddfc16cb4 100644 (file)
--- a/main.c
+++ b/main.c
@@ -66,7 +66,7 @@ void *worker(void *arg) {
         rc = db_insert(thrinf->db, entry, cfg.rescan, cfg.blur);
         if ( 0 != rc ) {
             if ( 0 > rc )
-                eprintf("not fingerprinting '%s'\n", entry->fname);
+                eprintf("WARNING: not fingerprinting '%s'\n", entry->fname);
             s_free(entry->fname);
             s_free(entry);
         }
@@ -111,7 +111,7 @@ int scan_dir(const char *dir, queue_t *q, db_t *db) {
         s_free(dirpath);
     }
     else
-        eprintf("%s: '%s'\n", strerror(errno), dir);
+        eprintf("ERROR: '%s': %s\n", dir, strerror(errno));
     /* wait for workers to finish */
     q_set_complete(q);
     for ( int i = 0; i < cfg.nthreads; ++i )
@@ -139,7 +139,7 @@ static int find_dupes_cb(db_entry_t *dupes) {
 static void usage(char *pname, int ec) {
     char *prog = basename(pname);
     printf("%s - find potentially duplicate images\n", prog);
-    printf("USAGE:  %s [OPTIONS] DIR ...\n", prog);
+    printf("USAGE:  %s [OPTIONS] [DIR ...]\n", prog);
     printf("OPTIONS:\n"
            "  -d n      max directory recursion depth; default: 0 (unlimited)\n"
            "  -f file   database file\n"
@@ -157,7 +157,8 @@ static void usage(char *pname, int ec) {
 
 /* main function */
 int main(int argc, char *argv[]) {
-    int c, rc = 0, n;
+    int c, rc = 0, n, err = 0;
+    char **dirs = NULL;
     char *cp;
     db_t *db ;
     const char *lead_in =
@@ -225,43 +226,42 @@ int main(int argc, char *argv[]) {
     }
 
     if ( cfg.db_prune ) {
-        dprintf("pruning\n");
+        dprintf("pruning missing files from database\n");
         db_prune(db);
     }
 
     q = q_init();
-    if (optind >= argc) {
-        dprintf("scanning '.'\n");
-        rc = scan_dir(".", q, db);
-    } else {
-        while ( optind < argc ){//&& 0 == rc ) {
-            dprintf("scanning '%s'\n", argv[optind]);
-            q_reset_complete(q);
-            rc = scan_dir(argv[optind++], q, db);
+    dirs = optind < argc ? &argv[optind] : (char *[]){".", NULL};
+    for ( int i = 0; NULL != dirs[i]; ++i ) {
+        dprintf("scanning '%s'\n", dirs[i]);
+        q_reset_complete(q);
+        rc = scan_dir(dirs[i], q, db);
+        if ( 0 != rc ) {
+            eprintf("ERROR: scanning '%s' failed\n", dirs[i]);
+            ++err;
         }
     }
     q_destroy(&q);
-    if ( 0 != rc )
-        goto DONE;
 
     if ( NULL != cfg.db_outfile ) {
         int cnt;
         cnt = db_write(db, cfg.db_outfile);
         if ( 0 <= cnt )
             dprintf("%d entries written\n", cnt);
-        else
-            eprintf("writing '%s' failed\n", cfg.db_outfile);
+        else {
+            eprintf("ERROR: writing '%s' failed\n", cfg.db_outfile);
+            ++err;
+        }
     }
 
-    dprintf("searching dupes ...\n");
+    dprintf("searching for potential duplicates ...\n");
     printf("%s", lead_in);
-    rc = db_find_dupes(db, cfg.thresh, find_dupes_cb);
+    db_find_dupes(db, cfg.thresh, find_dupes_cb);
     printf("END\n");
 
-  DONE:
     db_destroy(&db);
-    dprintf("done.\n");
-    exit(rc ? EXIT_FAILURE : EXIT_SUCCESS);
+    dprintf("done, encountered %d error%s\n", err, err==1?"":"s");
+    exit(err ? EXIT_FAILURE : EXIT_SUCCESS);
 }
 
 /* EOF */