fix cache key

CHANGELOG.md

@@ -1,6 +1,7 @@
 # 0.9.1 (2019-06-16)
 
 -   Add enabled adapters to cache key if caching for archive
+-   Prevent empty trailing page output in pdf reader
 
 # 0.9.0 (2019-06-16)
 

src/adapters.rs

@@ -26,7 +26,9 @@ pub struct AdapterMeta {
     /// version identifier. used to key cache entries, change if your output format changes
     pub version: i32,
     pub description: String,
-    /// list of matchers (interpreted as ORed)
+    /// indicates whether this adapter can descend (=call rga_preproc again). if true, the cache key needs to include the list of active adapters
+    pub recurses: bool,
+    /// list of matchers (interpreted as a OR b OR ...)
     pub fast_matchers: Vec<FastMatcher>,
     /// list of matchers when we have mime type detection active (interpreted as ORed)
     /// warning: this *overrides* the fast matchers
@@ -71,7 +73,6 @@ pub struct AdaptInfo<'a> {
     pub oup: &'a mut (dyn Write + Send),
     /// prefix every output line with this string to better indicate the file's location if it is in some archive
     pub line_prefix: &'a str,
-    // pub adapt_subobject: &'a dyn Fn(AdaptInfo) -> Fallible<()>,
     pub config: PreprocConfig<'a>,
 }
 

src/adapters/decompress.rs

@@ -19,6 +19,7 @@ lazy_static! {
         description:
             "Reads compressed file as a stream and runs a different extractor on the contents."
                 .to_owned(),
+        recurses: true,
         fast_matchers: EXTENSIONS
             .iter()
             .map(|s| FastMatcher::FileExtension(s.to_string()))

src/adapters/ffmpeg.rs

@@ -16,6 +16,7 @@ lazy_static! {
         name: "ffmpeg".to_owned(),
         version: 1,
         description: "Uses ffmpeg to extract video metadata/chapters and subtitles".to_owned(),
+        recurses: false,
         fast_matchers: EXTENSIONS
             .iter()
             .map(|s| FastMatcher::FileExtension(s.to_string()))
@@ -47,7 +48,7 @@ struct FFprobeStream {
     codec_type: String, // video,audio,subtitle
 }
 impl FileAdapter for FFmpegAdapter {
-    fn adapt(&self, ai: AdaptInfo, detection_reason: &SlowMatcher) -> Fallible<()> {
+    fn adapt(&self, ai: AdaptInfo, _detection_reason: &SlowMatcher) -> Fallible<()> {
         let AdaptInfo {
             is_real_file,
             filepath_hint,

src/adapters/pandoc.rs

@@ -48,6 +48,7 @@ lazy_static! {
         description:
             "Uses pandoc to convert binary/unreadable text documents to plain markdown-like text"
                 .to_owned(),
+        recurses: false,
         fast_matchers: EXTENSIONS
             .iter()
             .map(|s| FastMatcher::FileExtension(s.to_string()))

src/adapters/pdfpages.rs

@@ -17,6 +17,7 @@ lazy_static! {
 		name: "pdfpages".to_owned(),
 		version: 1,
 		description: "Converts a pdf to it's individual pages as png files. Only useful in combination with tesseract".to_owned(),
+		recurses: true,
 		fast_matchers: EXTENSIONS
 			.iter()
 			.map(|s| FastMatcher::FileExtension(s.to_string()))
@@ -42,7 +43,7 @@ impl GetMetadata for PdfPagesAdapter {
 /// A pdf is basically converted to a zip that has Page X.png files.
 /// This way, something like tesseract can process the pages individually
 impl FileAdapter for PdfPagesAdapter {
-	fn adapt(&self, ai: AdaptInfo, detection_reason: &SlowMatcher) -> Fallible<()> {
+	fn adapt(&self, ai: AdaptInfo, _detection_reason: &SlowMatcher) -> Fallible<()> {
 		let AdaptInfo {
 			filepath_hint,
 			is_real_file,

src/adapters/poppler.rs

@@ -12,6 +12,7 @@ lazy_static! {
         version: 1,
         description: "Uses pdftotext (from poppler-utils) to extract plain text from PDF files"
             .to_owned(),
+        recurses: false,
         fast_matchers: EXTENSIONS
             .iter()
             .map(|s| FastMatcher::FileExtension(s.to_string()))
@@ -50,6 +51,9 @@ impl SpawningFileAdapter for PopplerAdapter {
                 // page break
                 line = line.replace('\x0c', "");
                 page += 1;
+                if line.is_empty() {
+                    continue;
+                }
             }
             oup.write_all(format!("{}Page {}: {}\n", line_prefix, page, line).as_bytes())?;
         }

src/adapters/spawning.rs

@@ -93,7 +93,7 @@ impl<T> FileAdapter for T
 where
     T: SpawningFileAdapter,
 {
-    fn adapt(&self, ai: AdaptInfo, detection_reason: &SlowMatcher) -> Fallible<()> {
+    fn adapt(&self, ai: AdaptInfo, _detection_reason: &SlowMatcher) -> Fallible<()> {
         let AdaptInfo {
             filepath_hint,
             mut inp,

src/adapters/sqlite.rs

@@ -14,6 +14,7 @@ lazy_static! {
         description:
             "Uses sqlite bindings to convert sqlite databases into a simple plain text format"
                 .to_owned(),
+        recurses: false, // set to true if we decide to make sqlite blobs searchable (gz blob in db is kinda common I think)
         fast_matchers: EXTENSIONS
             .iter()
             .map(|s| FastMatcher::FileExtension(s.to_string()))
@@ -56,7 +57,7 @@ fn format_blob(b: ValueRef) -> String {
 }
 
 impl FileAdapter for SqliteAdapter {
-    fn adapt(&self, ai: AdaptInfo, detection_reason: &SlowMatcher) -> Fallible<()> {
+    fn adapt(&self, ai: AdaptInfo, _detection_reason: &SlowMatcher) -> Fallible<()> {
         let AdaptInfo {
             is_real_file,
             filepath_hint,

src/adapters/tar.rs

@@ -13,6 +13,7 @@ lazy_static! {
         name: "tar".to_owned(),
         version: 1,
         description: "Reads a tar file as a stream and recurses down into its contents".to_owned(),
+        recurses: true,
         fast_matchers: EXTENSIONS
             .iter()
             .map(|s| FastMatcher::FileExtension(s.to_string()))
@@ -35,7 +36,7 @@ impl GetMetadata for TarAdapter {
 }
 
 impl FileAdapter for TarAdapter {
-    fn adapt(&self, ai: AdaptInfo, detection_reason: &SlowMatcher) -> Fallible<()> {
+    fn adapt(&self, ai: AdaptInfo, _detection_reason: &SlowMatcher) -> Fallible<()> {
         let AdaptInfo {
             filepath_hint,
             mut inp,

src/adapters/tesseract.rs

@@ -10,6 +10,7 @@ lazy_static! {
 		name: "tesseract".to_owned(),
 		version: 1,
 		description: "Uses tesseract to run OCR on images to make them searchable. May need -j1 to prevent overloading the system. Make sure you have tesseract installed.".to_owned(),
+		recurses: false,
 		fast_matchers: EXTENSIONS
 			.iter()
 			.map(|s| FastMatcher::FileExtension(s.to_string()))

src/adapters/zip.rs

@@ -14,6 +14,7 @@ lazy_static! {
         name: "zip".to_owned(),
         version: 1,
         description: "Reads a zip file as a stream and recurses down into its contents".to_owned(),
+        recurses: true,
         fast_matchers: EXTENSIONS
             .iter()
             .map(|s| FastMatcher::FileExtension(s.to_string()))
@@ -45,7 +46,7 @@ fn is_dir(f: &ZipFile) -> bool {
 }
 
 impl FileAdapter for ZipAdapter {
-    fn adapt(&self, ai: AdaptInfo, detection_reason: &SlowMatcher) -> Fallible<()> {
+    fn adapt(&self, ai: AdaptInfo, _detection_reason: &SlowMatcher) -> Fallible<()> {
         let AdaptInfo {
             filepath_hint,
             mut inp,

src/preproc.rs

@@ -72,13 +72,22 @@ pub fn rga_preproc(ai: AdaptInfo) -> Result<(), Error> {
                     let clean_path = filepath_hint.to_owned().clean();
                     let meta = std::fs::metadata(&filepath_hint)?;
 
-                    let key = (
-                        clean_path,
-                        meta.modified().expect("weird OS that can't into mtime"),
-                    );
-                    eprintln!("cache key: {:?}", key);
-
-                    bincode::serialize(&key).expect("could not serialize path") // key in the cache database
+                    if adapter.metadata().recurses {
+                        let key = (
+                            clean_path,
+                            meta.modified().expect("weird OS that can't into mtime"),
+                            &args.adapters[..],
+                        );
+                        eprintln!("cache key: {:?}", key);
+                        bincode::serialize(&key).expect("could not serialize path") // key in the cache database
+                    } else {
+                        let key = (
+                            clean_path,
+                            meta.modified().expect("weird OS that can't into mtime"),
+                        );
+                        eprintln!("cache key: {:?}", key);
+                        bincode::serialize(&key).expect("could not serialize path") // key in the cache database
+                    }
                 };
                 cache.write().unwrap().get_or_run(
                     &db_name,