Merge pull request #628 from fubar2/fastaCLpulsarfix

fasta_compute_length: Add required file for Pulsar deployments and bump python to 3.12

tools/fasta_compute_length/.shed.yml

@@ -9,4 +9,5 @@ long_description: |
 name: fasta_compute_length
 owner: devteam
 remote_repository_url: https://github.com/galaxyproject/tools-devteam/tree/master/tools/fasta_compute_length
+homepage_url: https://github.com/galaxyproject/tools-devteam/tree/master/tools/fasta_compute_length
 type: unrestricted

tools/fasta_compute_length/fasta_compute_length.py

@@ -1,10 +1,47 @@
 #!/usr/bin/env python
 """
-Uses fasta_to_len converter code.
+Input: fasta, int
+Output: tabular
+Return titles with lengths of corresponding seq
 """
 
 import sys
-from utils.fasta_to_len import compute_fasta_length
 
 
-compute_fasta_length(sys.argv[1], sys.argv[2], sys.argv[3], sys.argv[4] == 'id_only')
+def compute_fasta_length(fasta_file, out_file, keep_first_char, keep_first_word=False):
+    keep_first_char = int(keep_first_char)
+    fasta_title = ''
+    seq_len = 0
+
+    # number of char to keep in the title
+    if keep_first_char == 0:
+        keep_first_char = None
+    else:
+        keep_first_char += 1
+
+    first_entry = True
+    with open(fasta_file) as in_fh, open(out_file, 'w') as out_fh:
+        for line in in_fh:
+            line = line.strip()
+            if not line or line.startswith('#'):
+                continue
+            if line[0] == '>':
+                if first_entry is False:
+                    if keep_first_word:
+                        fasta_title = fasta_title.split()[0]
+                    out_fh.write("%s\t%d\n" % (fasta_title[1:keep_first_char], seq_len))
+                else:
+                    first_entry = False
+                fasta_title = line
+                seq_len = 0
+            else:
+                seq_len += len(line)
+
+        # last fasta-entry
+        if keep_first_word:
+            fasta_title = fasta_title.split()[0]
+        out_fh.write("%s\t%d\n" % (fasta_title[1:keep_first_char], seq_len))
+
+
+if __name__ == "__main__":
+    compute_fasta_length(sys.argv[1], sys.argv[2], sys.argv[3], sys.argv[4] == 'id_only')

tools/fasta_compute_length/fasta_compute_length.xml

@@ -1,8 +1,11 @@
-<tool id="fasta_compute_length" name="Compute sequence length" version="1.0.3" profile="16.04">
+<tool id="fasta_compute_length" name="Compute sequence length" version="1.0.4" profile="22.04">
     <description></description>
     <requirements>
-        <requirement type="package" version="3.7">python</requirement>
+        <requirement type="package" version="3.12">python</requirement>
     </requirements>
+    <required_files>
+        <include path="fasta_compute_length.py"/>
+    </required_files>
     <command>
     #if $ref.ref_source == 'dbkey':
         cp '${ref.index.fields.len_path}' '$output'