fix recursion and pdf adapter

src/adapters.rs

@@ -25,6 +25,8 @@ use std::path::PathBuf;
 use std::pin::Pin;
 use std::rc::Rc;
 
+use self::postproc::PostprocPageBreaks;
+
 pub type ReadBox = Pin<Box<dyn AsyncRead + Send>>;
 pub struct AdapterMeta {
     /// unique short name of this adapter (a-z0-9 only)
@@ -117,6 +119,7 @@ pub fn get_all_adapters(custom_adapters: Option<Vec<CustomAdapterConfig>>) -> Ad
     }
 
     let internal_adapters: Vec<Arc<dyn FileAdapter>> = vec![
+        Arc::new(PostprocPageBreaks::new()),
         //Rc::new(ffmpeg::FFmpegAdapter::new()),
         // Rc::new(zip::ZipAdapter::new()),
         //Rc::new(decompress::DecompressAdapter::new()),

src/adapters/custom.rs

@@ -213,9 +213,6 @@ fn arg_replacer(arg: &str, filepath_hint: &Path) -> Result<String> {
     }))
 }
 impl CustomSpawningFileAdapter {
-    fn get_exe(&self) -> &str {
-        &self.binary
-    }
     fn command(
         &self,
         filepath_hint: &std::path::Path,
@@ -266,7 +263,7 @@ impl FileAdapter for CustomSpawningFileAdapter {
             inp: output,
             line_prefix,
             is_real_file: false,
-            archive_recursion_depth,
+            archive_recursion_depth: archive_recursion_depth + 1,
             postprocess,
             config,
         })))

src/adapters/postproc.rs

@@ -8,7 +8,9 @@ use async_stream::stream;
 use bytes::Bytes;
 use encoding_rs_io::DecodeReaderBytesBuilder;
 use std::cmp::min;
+use std::ffi::OsStr;
 use std::io::Cursor;
+use std::path::PathBuf;
 use std::pin::Pin;
 use tokio::io::{AsyncRead, AsyncReadExt};
 use tokio_util::io::ReaderStream;
@@ -145,6 +147,11 @@ pub fn postproc_prefix(line_prefix: &str, inp: impl AsyncRead + Send) -> impl As
 }
 
 pub struct PostprocPageBreaks {}
+impl PostprocPageBreaks {
+    pub fn new() -> Self {
+        Self {}
+    }
+}
 impl GetMetadata for PostprocPageBreaks {
     fn metadata(&self) -> &super::AdapterMeta {
         lazy_static::lazy_static! {
@@ -176,6 +183,13 @@ impl FileAdapter for PostprocPageBreaks {
         let ai = AdaptInfo {
             inp: Box::pin(read),
             postprocess: false,
+            archive_recursion_depth: a.archive_recursion_depth + 1,
+            filepath_hint: a
+                .filepath_hint
+                .parent()
+                .map(PathBuf::from)
+                .unwrap_or(PathBuf::new())
+                .join(a.filepath_hint.file_stem().unwrap_or(OsStr::new(""))),
             ..a
         };
         Ok(Box::pin(tokio_stream::once(ai)))

src/preproc.rs

@@ -189,8 +189,7 @@ async fn adapt_caching(
     match cached {
         Some(cached) => Ok(Box::pin(ZstdDecoder::new(Cursor::new(cached)))),
         None => {
-            debug!("cache MISS, running adapter");
-            debug!("adapting with caching...");
+            debug!("cache MISS, running adapter with caching...");
             let inp = loop_adapt(adapter.as_ref(), detection_reason, ai)?;
             let inp = concat_read_streams(inp);
             let inp = async_read_and_write_to_cache(
@@ -228,16 +227,27 @@ pub fn loop_adapt(
             adapter.metadata().name
         )
     })?;
+    debug!("got fph starting loop: {}", fph.to_string_lossy());
 
     let s = stream! {
         for await file in inp {
             match buf_choose_adapter(file).await.expect("todo: handle") {
-                Ret::Recurse(ai, adapter, detection_reason, active_adapters) => {
+                Ret::Recurse(ai, adapter, detection_reason, _active_adapters) => {
+                    debug!(
+                        "Chose adapter '{}' because of matcher {:?}",
+                        &adapter.metadata().name, &detection_reason
+                    );
+                    eprintln!(
+                        "{} adapter: {}",
+                        ai.filepath_hint.to_string_lossy(),
+                        &adapter.metadata().name
+                    );
                     for await ifile in loop_adapt(adapter.as_ref(), detection_reason, ai).expect("todo: handle") {
                         yield ifile;
                     }
                 }
                 Ret::Passthrough(ai) => {
+                    debug!("no adapter for {}, ending recursion", ai.filepath_hint.to_string_lossy());
                     yield ai;
                 }
             }