From bca435557b579f6d1c43a72a35f2fcf87b3979cb Mon Sep 17 00:00:00 2001
From: phiresky <phireskyde+git@gmail.com>
Date: Tue, 11 Jun 2019 14:11:08 +0200
Subject: [PATCH] fix --rga-accurate (matching by mime types)

---
 exampledir/sqlitedb    | Bin 0 -> 8192 bytes
 src/adapters/ffmpeg.rs |   1 +
 src/bin/rga-preproc.rs |   6 +++---
 src/bin/rga.rs         |  25 ++++++++++++++-----------
 src/preproc.rs         |  21 +++++++++++++++------
 5 files changed, 33 insertions(+), 20 deletions(-)
 create mode 100644 exampledir/sqlitedb

diff --git a/exampledir/sqlitedb b/exampledir/sqlitedb
new file mode 100644
index 0000000000000000000000000000000000000000..efb082b2f92776bbc45b3576860910be50d4a67b
GIT binary patch
literal 8192
zcmeI#u?oU45C-78L<9$Ep+nbzg9_rJ;A)j_E^1#Og_H`FK&goMg1&(d;4`>;9Xh+X
z2>;Jrl7md&cE|nh)HqZ4`fMAsppckpZ441<X(io-<<aW~x!d_$G|iW0D8D#(k`w{~
z2tWV=5P$##AOHafKmY;|fWYqr#@vsih*yoxD_3|grM8DO$x@Rgn_Vgkzn|DyK5ch%
z6}%{T%Yi@u0uX=z1Rwwb2tWV=5P$##An>mOq0%E&`m3?EGdF+X$AERrrK>7`{{T2R
B9uWWl

literal 0
HcmV?d00001

diff --git a/src/adapters/ffmpeg.rs b/src/adapters/ffmpeg.rs
index efc5cbf..dd0e4d3 100644
--- a/src/adapters/ffmpeg.rs
+++ b/src/adapters/ffmpeg.rs
@@ -125,6 +125,7 @@ impl FileAdapter for FFmpegAdapter {
             let stdo = cmd.stdout.as_mut().expect("is piped");
             let time_re = Regex::new(r".*\d.*-->.*\d.*").unwrap();
             let mut time: String = "".to_owned();
+            // rewrite subtitle times so they are prefixed in every line
             for line in BufReader::new(stdo).lines() {
                 let line = line?;
                 // 09:55.195 --> 09:56.730
diff --git a/src/bin/rga-preproc.rs b/src/bin/rga-preproc.rs
index 690ddfc..f1b9356 100644
--- a/src/bin/rga-preproc.rs
+++ b/src/bin/rga-preproc.rs
@@ -4,7 +4,7 @@ use rga::adapters::*;
 use rga::preproc::*;
 
 use std::fs::File;
-use std::io::BufReader;
+
 fn main() -> Fallible<()> {
     env_logger::init();
     let empty: Vec<std::ffi::OsString> = vec![];
@@ -19,7 +19,7 @@ fn main() -> Fallible<()> {
         std::env::current_dir()?.join(&filepath)
     };
 
-    let i = File::open(&path)?;
+    let mut i = File::open(&path)?;
     let mut o = std::io::stdout();
     let cache = if args.no_cache {
         None
@@ -27,7 +27,7 @@ fn main() -> Fallible<()> {
         Some(rga::preproc_cache::open()?)
     };
     let ai = AdaptInfo {
-        inp: &mut BufReader::new(i),
+        inp: &mut i,
         filepath_hint: &path,
         is_real_file: true,
         oup: &mut o,
diff --git a/src/bin/rga.rs b/src/bin/rga.rs
index 15c09a1..a9bdaf9 100644
--- a/src/bin/rga.rs
+++ b/src/bin/rga.rs
@@ -4,10 +4,8 @@ use rga::adapters::spawning::map_exe_error;
 use rga::adapters::*;
 use rga::args::*;
 
-
 use std::process::Command;
 
-
 fn main() -> Fallible<()> {
     env_logger::init();
 
@@ -35,14 +33,19 @@ fn main() -> Fallible<()> {
         return Ok(());
     }
 
-    let extensions = adapters
-        .iter()
-        .flat_map(|a| &a.metadata().fast_matchers)
-        .filter_map(|m| match m {
-            FastMatcher::FileExtension(ext) => Some(ext as &str),
-        })
-        .collect::<Vec<_>>()
-        .join(",");
+    let pre_glob = if !args.accurate {
+        let extensions = adapters
+            .iter()
+            .flat_map(|a| &a.metadata().fast_matchers)
+            .filter_map(|m| match m {
+                FastMatcher::FileExtension(ext) => Some(ext as &str),
+            })
+            .collect::<Vec<_>>()
+            .join(",");
+        format!("*.{{{}}}", extensions)
+    } else {
+        "*".to_owned()
+    };
 
     let exe = std::env::current_exe().expect("Could not get executable location");
     let preproc_exe = exe.with_file_name("rga-preproc");
@@ -51,7 +54,7 @@ fn main() -> Fallible<()> {
         .arg("--pre")
         .arg(preproc_exe)
         .arg("--pre-glob")
-        .arg(format!("*.{{{}}}", extensions))
+        .arg(pre_glob)
         .args(passthrough_args)
         .spawn()
         .map_err(|e| map_exe_error(e, "rg", "Please make sure you have ripgrep installed."))?;
diff --git a/src/preproc.rs b/src/preproc.rs
index 0af12e1..5fb5880 100644
--- a/src/preproc.rs
+++ b/src/preproc.rs
@@ -5,6 +5,8 @@ use failure::Fallible;
 use failure::{format_err, Error};
 use path_clean::PathClean;
 use std::convert::TryInto;
+use std::io::BufRead;
+use std::io::BufReader;
 use std::io::BufWriter;
 use std::sync::{Arc, RwLock};
 
@@ -43,13 +45,20 @@ pub fn rga_preproc(ai: AdaptInfo) -> Result<(), Error> {
 
     eprintln!("path_hint: {:?}", filepath_hint);
 
-    /*let mimetype = tree_magic::from_filepath(path).ok_or(lerr(format!(
-        "File {} does not exist",
-        filename.to_string_lossy()
-    )))?;
-    println!("mimetype: {:?}", mimetype);*/
+    // todo: figure out when using a bufreader is a good idea and when it is not
+    // seems to beed for File::open() reads, but not sure about within archives (tar, zip)
+    let inp = &mut BufReader::with_capacity(1 << 13, inp);
+
+    let mimetype = if args.accurate {
+        let buf = inp.fill_buf()?; // fill but do not consume!
+        let mimetype = tree_magic::from_u8(buf);
+        eprintln!("mimetype: {:?}", mimetype);
+        Some(mimetype)
+    } else {
+        None
+    };
     let adapter = adapters(FileMeta {
-        mimetype: None,
+        mimetype,
         lossy_filename: filename.to_string_lossy().to_string(),
     });
     match adapter {