Merge pull request #35 from 4dn-dcic/dev2

Dev2

.travis.yml

@@ -1,4 +1,3 @@
-sudo: false
 language: r
 cache: packages
 
@@ -11,10 +10,10 @@ r:
   - devel
   - oldrel
 
-bioc_packages: 
-  - GenomicRanges
-  - InteractionSet
+bioc_packages: GenomicRanges
+bioc_packages: InteractionSet
 
 # Upload CodeCov information
 after_success:
   - Rscript -e 'library(covr);codecov()'
+

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: Rpairix
 Title: Rpairix
-Version: 0.2.5
+Version: 0.3.5
 Authors@R: person("Soo", "Lee", email = "duplexa@gmail.com", role = c("aut", "cre"))
 Description: R binder for pairix, tool for querying a pair of genomic ranges in a pairs file (pairix-indexed bgzipped text file)
 Depends:
@@ -12,4 +12,6 @@ License: MIT
 Encoding: UTF-8
 LazyData: true
 RoxygenNote: 6.0.1
-Suggests: testthat
+Suggests:
+    testthat
+

R/px_query.R

@@ -94,6 +94,7 @@
 #' print(res)
 #' 
 #' # query with GRangesList
+#' class(grl)
 #' res = px_query(filename,query=grl)
 #' print(res)
 #'
@@ -117,7 +118,7 @@ px_query<-function(filename, query, max_mem=100000000, stringsAsFactors=FALSE, l
     qdf <- qdf[,c("seqnames1","start1","end1","seqnames2","start2","end2")]
     querystr <- df_to_querystr(qdf)
     rm(qdf)
-  } else if (class(query)=="GRangesList"){
+  } else if (class(query)=="GRangesList" || class(query)=="CompressedGRangesList"){
     # -- produce querystr from two identical-length, paired GRanges objects in a GRangesList-- #
     # test lengths
     if(length(query) != 2) stop("GRangesList must be composed of two GRanges objects.")

README.md

@@ -55,7 +55,7 @@ R --no-site-file --no-environ --no-save --no-restore CMD INSTALL --install-tests
 To install a specific version,
 ```r
 library(devtools)
-install_url("https://github.com/4dn-dcic/Rpairix/archive/0.2.5.zip")
+install_url("https://github.com/4dn-dcic/Rpairix/archive/0.3.5.zip")
 ```
 
 
@@ -325,6 +325,13 @@ Individual R functions are written and documented in `R/`. The `src/rpairixlib.c
 ***
 
 ## Version history
+### 0.3.5
+* Index structure and C source codes are consistent with pairix/pypairix 0.3.5. This new structure can deal with large chromosomes (>length 2^29). The older index can be read and used for regular chromosomes (<2^29) (backward-compatible).
+
+### 0.2.5
+* Index structure and C source codes are consistent with pairix/pypairix 0.2.5 (please re-index if your index was built by an older version of pairix/pypairix/Rpairix)
+* `px_build_index` function now has parameter `region_split_character` (default: '|'). Query must use the same region_split_character used for building the index.
+
 ### 0.2.4
 * Index is compatible with pairix/pypairix 0.2.4. (Magic number updated)
 

Rpairix.Rproj

@@ -1,4 +1,4 @@
-Version: 0.2.5
+Version: 0.3.5
 
 RestoreWorkspace: No
 SaveWorkspace: No

src/bgzf.c

@@ -325,6 +325,27 @@ static int load_block_from_cache(BGZF *fp, int64_t block_address) {return 0;}
 static void cache_block(BGZF *fp, int size) {}
 #endif
 
+int bgzf_block_length(BGZF *fp, int64_t block_start_offset)
+{
+	uint8_t header[BLOCK_HEADER_LENGTH], *compressed_block;
+	int count, block_length, remaining;
+	int64_t block_address;
+        bgzf_seek(fp, block_start_offset, SEEK_SET);
+	block_address = _bgzf_tell((_bgzf_file_t)fp->fp);
+	if (load_block_from_cache(fp, block_address)) return 0;
+	count = _bgzf_read(fp->fp, header, sizeof(header));
+	if (count == 0) { // no data read
+		fp->block_length = 0;
+		return 0;
+	}
+	if (count != sizeof(header) || !check_header(header)) {
+		fp->errcode |= BGZF_ERR_HEADER;
+		return -1;
+	}
+	block_length = unpackInt16((uint8_t*)&header[16]) + 1; // +1 because when writing this number, we used "-1"
+        return(block_length);
+}
+
 int bgzf_read_block(BGZF *fp)
 {
 	uint8_t header[BLOCK_HEADER_LENGTH], *compressed_block;

src/bgzf.h

@@ -186,6 +186,11 @@ extern "C" {
 	 */
 	int bgzf_read_block(BGZF *fp);
 
+        /**
+         * returns block length for a given block start offset
+         */
+        int bgzf_block_length(BGZF *fp, int64_t block_start_offset);
+
 #ifdef __cplusplus
 }
 #endif