Eliminated non-text datatypes from sniff order since galaxy hinders non-text file uploads (e.g., gmaj.zip files are sniffed as txt data types since Galaxy automaticaally changes EOL

characters upon upload. Fixed the guess_ext() function in sniff.py to eliminate the call to sep2tabs since this function changes the data in the file (the guess_ext() function probably shouldn't be doing that). Added a sniff function to the Gmaj class as an example for zip data types.
2026-09-21 13:50:20 +08:00 · 2007-10-01 21:06:58 +00:00
parent e73afecfde
commit ff7ecaa58d
5 changed files with 53 additions and 37 deletions
@@ -6,6 +6,7 @@ import data
 import logging
 from galaxy.datatypes.sniff import *
 from urllib import urlencode
+import zipfile

 log = logging.getLogger(__name__)

@@ -15,7 +16,6 @@ class Image( data.Data ):
        dataset.peek  = 'Image in %s format (%s)' % ( dataset.extension, data.nice_size( dataset.get_size() ) )
        dataset.blurb = 'image' 

-
 class Gmaj( data.Data ):
    """Class describing a GMAJ Applet"""
    file_ext = "gmaj.zip"
@@ -32,10 +32,25 @@ class Gmaj( data.Data ):
    def get_mime(self):
        """Returns the mime type of the datatype"""
        return 'application/zip'
-    def sniff( self, filename ):
-        #TODO: fix me
-        return False
-
+    def sniff(self, filename):
+        """
+        NOTE: the sniff.convert_newlines() call in the upload utility will keep Gmaj data types from being 
+        correctly sniffed, but the files can be uploaded (they'll be sniffed as 'txt').  This sniff function
+        is here to provide an example of a sniffer for a zip file.
+        """
+        if not zipfile.is_zipfile( filename ):
+            return False
+        contains_gmaj_file = False
+        zip_file = zipfile.ZipFile(filename, "r")
+        for name in zip_file.namelist():
+            if name.split(".")[1].strip().lower() == 'gmaj':
+                contains_gmaj_file = True
+                break
+        zip_file.close()
+        if not contains_gmaj_file:
+            return False
+        return True
+            
 class Html( data.Text ):
    """Class describing an html file"""
    file_ext = "html"
@@ -60,7 +75,6 @@ class Html( data.Text ):
        True
        """
        headers = get_headers( filename, None )
-
        try:
            for i, hdr in enumerate(headers):
                if hdr and hdr[0].lower().find( '<html>' ) >=0:
@@ -82,7 +96,4 @@ class Laj( data.Text ):
            return dataset.peek
        except:
            return "peek unavailable"
-    def sniff( self, filename ):
-        #TODO: fix me...
-        return False

@@ -102,8 +102,6 @@ class Registry( object ):
        #default values
        if len(self.sniff_order) < 1:
            self.sniff_order = [
-                images.Gmaj(),
-                images.Laj(),
                sequence.Maf(),
                sequence.Lav(),
                sequence.Fasta(),
@@ -161,16 +161,19 @@ def guess_ext( fname ):
    >>> guess_ext(fname)
    'gff3'
    >>> fname = get_test_fname('temp.txt')
-    >>> file(fname, 'wt').write("a 2\\nc 1\\nd 0")
+    >>> file(fname, 'wt').write("a\\t2\\nc\\t1\\nd\\t0")
    >>> guess_ext(fname)
    'tabular'
    >>> fname = get_test_fname('temp.txt')
    >>> file(fname, 'wt').write("a 1 2 x\\nb 3 4 y\\nc 5 6 z")
    >>> guess_ext(fname)
-    'tabular'
+    'txt'
    >>> fname = get_test_fname('test_tab1.tabular')
    >>> guess_ext(fname)
    'tabular'
+    >>> fname = get_test_fname('alignment.lav')
+    >>> guess_ext(fname)
+    'lav'
    """ 
    datatypes_registry = registry.Registry()
    for datatype in datatypes_registry.sniff_order:
@@ -188,19 +191,21 @@ def guess_ext( fname ):
        except:
            pass

-    """Default binary file extension"""
-    for line in file( fname ):
-        for char in line:
-            if ord(char) > 128:
-                return 'data'
-            else:
+    headers = get_headers( fname, None )
+    is_binary = True
+    for hdr in headers:
+        for char in hdr:
+            try:
+                if not ord(char) > 128:
+                    is_binary = False
+            except:
+                is_binary = False
                break
-        break
-    if is_column_based( fname, ' ', 1 ):
-        sep2tabs(fname)
+    if is_binary:
+        return 'data'        #default binary data type file extension
    if is_column_based( fname, '\t', 1):
-        return 'tabular'
-    return 'txt'
+        return 'tabular'    #default tabular data type file extension
+    return 'txt'            #default text data type file extension

 if __name__ == '__main__':
    import doctest, sys
@@ -87,6 +87,10 @@ class UploadToolAction( object ):
        if self.empty:
            raise BadFileException( "attempted to upload an empty or inappropriate file" )
        
+        """
+        NOTE: the following will keep binary and zip files (e.g., gmaj.zip) from being correctly sniffed, but
+        the files can be uploaded (they'll be sniffed as 'txt').  This should restrict some unwanted behavior.
+        """
        sniff.convert_newlines(temp_name)
        
        if space_to_tab:
@@ -174,16 +174,14 @@ nametable = galaxy.datatypes.data:Text

 [galaxy:sniff_order]

-05 = galaxy.datatypes.images:Gmaj
-10 = galaxy.datatypes.images:Laj
-15 = galaxy.datatypes.sequence:Maf
-20 = galaxy.datatypes.sequence:Lav
-25 = galaxy.datatypes.sequence:Fasta
-30 = galaxy.datatypes.interval:Wiggle
-35 = galaxy.datatypes.images:Html
-40 = galaxy.datatypes.sequence:Axt
-45 = galaxy.datatypes.interval:Bed
-50 = galaxy.datatypes.interval:CustomTrack
-55 = galaxy.datatypes.interval:Gff
-60 = galaxy.datatypes.interval:Gff3
-65 = galaxy.datatypes.interval:Interval
+05 = galaxy.datatypes.sequence:Maf
+10 = galaxy.datatypes.sequence:Lav
+15 = galaxy.datatypes.sequence:Fasta
+20 = galaxy.datatypes.interval:Wiggle
+25 = galaxy.datatypes.images:Html
+30 = galaxy.datatypes.sequence:Axt
+35 = galaxy.datatypes.interval:Bed
+40 = galaxy.datatypes.interval:CustomTrack
+45 = galaxy.datatypes.interval:Gff
+50 = galaxy.datatypes.interval:Gff3
+55 = galaxy.datatypes.interval:Interval