Update readme

README.md

@@ -5,19 +5,40 @@
 
 sist2 (Simple incremental search tool)
 
+*Warning: sist2 is in early development*
+
 ## Features
 
+* Fast, low memory usage
+* Portable (all its features are packaged in a single executable)
+* Extracts text from common file types\*
+* Generates thumbnails\*
+* Incremental scanning
+
+
+\* See [format support](#format-support)
+
+## Getting Started
+
+1. Have an [Elasticsearch](https://www.elastic.co/downloads/elasticsearch) instance running
+1. Download the [latest sist2 release](https://github.com/simon987/sist2/releases)
+
+*Windows users*: `sist2` runs under [WSL](https://en.wikipedia.org/wiki/Windows_Subsystem_for_Linux)
+
+*Mac users*: See [#1](https://github.com/simon987/sist2/issues/1)
 
 
 ## Example usage
 
+![demo](demo.gif)
+
 See help page `sist2 --help` for more details.
 
 **Scan a directory**
 ```bash
 sist2 scan ~/Documents -o ./orig_idx/
 sist2 scan --threads 4 --content-size 16384 /mnt/Pictures
-sist2 scan -i ./orig_idx/ -o ./updated_idx/ ~/Documents
+sist2 scan --incremental ./orig_idx/ -o ./updated_idx/ ~/Documents
 ```
 
 **Push index to Elasticsearch or file**
@@ -46,22 +67,11 @@ docx, xlsx, pptx |  | *planned* | no | *planned* |
 
 
 
-## Installation
-
-1. Download runtime dependencies
-
-    `apt install curl bzip2`
-    
-1. Download binary
-    
-    Get [the latest release](https://github.com/simon987/sist2/releases) from GitHub
-
-1. (Optional) Add to search path
-    
-   `mv sist2 /usr/bin/`
-
 ## Build from source
 
+You can compile **sist2** by yourself if you don't want to use the pre-compiled
+binaries.
+
 1. Install compile-time dependencies
 
     *(Debian)*
@@ -70,12 +80,7 @@ docx, xlsx, pptx |  | *planned* | no | *planned* |
         libssl-dev uuid-dev libavformat-dev libswscale-dev \
         python3 libmagic-dev libfreetype6-dev libcurl-dev \
         libbz2-dev yasm
-    ```
+   
-    *(Archlinux)*
-    ```bash
-    pacman -S git ffmpeg pkg-config cmake openssl curl \
-        bzip2 yasm libutil-linux
-    ```
 2. Build
     ```bash
     git clone --recurse-submodules https://github.com/simon987/sist2

src/ctx.h

@@ -37,5 +37,4 @@ struct {
 } WebCtx;
 
 
-
 #endif

src/parsing/parse.c

@@ -94,7 +94,7 @@ void parse(void *arg) {
     if (!(SHOULD_PARSE(doc.mime))) {
 
     } else if ((mmime == MimeVideo && doc.size >= MIN_VIDEO_SIZE) || mmime == MimeAudio || mmime == MimeImage) {
-//        parse_media(job->filepath, &doc);
+        parse_media(job->filepath, &doc);
 
     } else if (IS_PDF(doc.mime)) {
         void *pdf_buf = read_all(job, (char *) buf, bytes_read, &fd);
@@ -105,15 +105,15 @@ void parse(void *arg) {
         }
 
     } else if (mmime == MimeText && ScanCtx.content_size > 0) {
-//        parse_text(bytes_read, &fd, (char *) buf, &doc);
+        parse_text(bytes_read, &fd, (char *) buf, &doc);
 
     } else if (IS_FONT(doc.mime)) {
-//        void *font_buf = read_all(job, (char *) buf, bytes_read, &fd);
+        void *font_buf = read_all(job, (char *) buf, bytes_read, &fd);
-//        parse_font(font_buf, doc.size, &doc);
+        parse_font(font_buf, doc.size, &doc);
-//
+
-//        if (font_buf != buf) {
+        if (font_buf != buf) {
-//            free(font_buf);
+            free(font_buf);
-//        }
+        }
     }
 
     write_document(&doc);

src/parsing/pdf.c

@@ -1,7 +1,7 @@
+#include <src/ctx.h>
 #include "pdf.h"
 #include "src/ctx.h"
 
-__always_inline
 fz_page *render_cover(fz_context *ctx, document_t *doc, fz_document *fzdoc) {
 
     fz_page *cover = fz_load_page(ctx, fzdoc, 0);
@@ -25,8 +25,12 @@ fz_page *render_cover(fz_context *ctx, document_t *doc, fz_document *fzdoc) {
     fz_device *dev = fz_new_draw_device(ctx, m, pixmap);
 
     pthread_mutex_lock(&ScanCtx.mupdf_mu);
+    fz_try(ctx)
         fz_run_page(ctx, cover, dev, fz_identity, NULL);
+    fz_always(ctx)
             pthread_mutex_unlock(&ScanCtx.mupdf_mu);
+    fz_catch(ctx)
+        fz_rethrow(ctx);
 
     fz_drop_device(ctx, dev);
 
@@ -72,6 +76,7 @@ void parse_pdf(void *buf, size_t buf_len, document_t *doc) {
         fzdoc = fz_open_document_with_stream(ctx, mime_get_mime_text(doc->mime), stream);
 
         int page_count = fz_count_pages(ctx, fzdoc);
+
         fz_page *cover = render_cover(ctx, doc, fzdoc);
 
         fz_stext_options opts;
@@ -90,8 +95,12 @@ void parse_pdf(void *buf, size_t buf_len, document_t *doc) {
             fz_device *dev = fz_new_stext_device(ctx, stext, &opts);
 
             pthread_mutex_lock(&ScanCtx.mupdf_mu);
+            fz_try(ctx)
                 fz_run_page_contents(ctx, page, dev, fz_identity, NULL);
+            fz_always(ctx)
                 pthread_mutex_unlock(&ScanCtx.mupdf_mu);
+            fz_catch(ctx)
+                fz_rethrow(ctx);
 
             fz_drop_device(ctx, dev);