Add zip_output (#240)

* Making zip of results * Zip command added, if zip is true place_report_in_output_folder is also true, zip removes all files while zipping * Adding --zip to compare and pooled/wgs compare * Add more formatting changes to CRISPRessoShared * Refactoring propagate_crispress_options so only one version exists * Zip added to arguments_to_ignore and warning added when changing arguments * Restore styling * Update README to include --zip * Rename --zip to --zip_output * Change --zip to --zip_output in CompareCORE and PooledWGSCompareCORE * Bug fix arg to args Co-authored-by: Samuel Nichols <Snic9004@gmail.com>
pinellolab · Aug 23, 2022 · c80f828 · c80f828
1 parent 5de3d72
commit c80f828
Show file tree

Hide file tree

Showing 8 changed files with 585 additions and 348 deletions.
diff --git a/CRISPResso2/CRISPRessoBatchCORE.py b/CRISPResso2/CRISPRessoBatchCORE.py
@@ -34,37 +34,6 @@
 _ROOT = os.path.abspath(os.path.dirname(__file__))
 
 ####Support functions###
-def propagate_options(cmd, options, params, paramInd):
-####
-# cmd - the command to run
-# options - list of options to propagate e.g. crispresso options
-# params - df from excel parser e.g. params['amplicon_name'] = ['name1','name2','name3'] where each item corresponds to a different run in the batch
-# paramInd - index in dict - this is the run number in the batch
-
-    for option in options:
-        if option:
-            if option in params:
-                val = params.loc[paramInd, option]
-                if val is None:
-                    pass
-                elif str(val) == "True":
-                    cmd += ' --%s' % option
-                elif str(val) == "False":
-                    pass
-                elif isinstance(val, str):
-                    if val != "":
-                        if " " in val or "-" in val:
-                            cmd += ' --%s "%s"' % (option, str(val))  # quotes for options with spaces
-                        else:
-                            cmd += ' --%s %s' % (option, str(val))
-                elif isinstance(val, bool):
-                    if val:
-                        cmd += ' --%s' % option
-                else:
-                    cmd += ' --%s %s' % (option, str(val))
-#    print("cmd is " + str(cmd))
-    return cmd
-
 def check_library(library_name):
     try:
         return __import__(library_name)
@@ -107,11 +76,15 @@ def main():
         debug_flag = args.debug
 
         crispresso_options = CRISPRessoShared.get_crispresso_options()
-        options_to_ignore = {'name', 'output_folder'}
+        options_to_ignore = {'name', 'output_folder', 'zip_output'}
         crispresso_options_for_batch = list(crispresso_options-options_to_ignore)
 
         CRISPRessoShared.check_file(args.batch_settings)
 
+        if args.zip_output and not args.place_report_in_output_folder:
+            logger.warn('Invalid arguement combination: If zip_output is True then place_report_in_output_folder must also be True. Setting place_report_in_output_folder to True.')
+            args.place_report_in_output_folder = True
+
         batch_folder_name = os.path.splitext(os.path.basename(args.batch_settings))[0]
         if args.name and args.name != "":
             clean_name = CRISPRessoShared.slugify(args.name)
@@ -298,7 +271,7 @@ def main():
             batch_input_names[batch_name] = row["name"]
 
             crispresso_cmd = args.crispresso_command + ' -o "%s" --name %s' % (OUTPUT_DIRECTORY, batch_name)
-            crispresso_cmd = propagate_options(crispresso_cmd, crispresso_options_for_batch, batch_params, idx)
+            crispresso_cmd = CRISPRessoShared.propagate_crispresso_options(crispresso_cmd, crispresso_options_for_batch, batch_params, idx)
             if row.amplicon_seq == "":
                 crispresso_cmd += ' --auto '
             crispresso_cmds.append(crispresso_cmd)
@@ -931,6 +904,12 @@ def main():
             crispresso2_info,
         )
         info('Analysis Complete!')
+        if args.zip_output:
+            if args.output_folder == "":
+                path_value = os.path.split(OUTPUT_DIRECTORY)
+                CRISPRessoShared.zip_results(path_value[1])
+            else:
+                CRISPRessoShared.zip_results(OUTPUT_DIRECTORY)
         print(CRISPRessoShared.get_crispresso_footer())
         sys.exit(0)
 

diff --git a/CRISPResso2/CRISPRessoCORE.py b/CRISPResso2/CRISPRessoCORE.py
@@ -975,6 +975,9 @@ def print_stacktrace_if_debug():
         CRISPRessoShared.check_file(aln_matrix_loc)
         aln_matrix = CRISPResso2Align.read_matrix(aln_matrix_loc)
 
+        if args.zip_output:
+            args.place_report_in_output_folder = True
+
         n_processes = 1
         if args.n_processes == "max":
             n_processes = CRISPRessoMultiProcessing.get_max_processes()
@@ -4816,6 +4819,8 @@ def get_scaffold_len(row, scaffold_start_loc, scaffold_seq):
             crispresso2_info_file,
             crispresso2_info,
         )
+        if args.zip_output:
+            CRISPRessoShared.zip_results(OUTPUT_DIRECTORY)
 
         info('Analysis Complete!')
         print(CRISPRessoShared.get_crispresso_footer())

diff --git a/CRISPResso2/CRISPRessoCompareCORE.py b/CRISPResso2/CRISPRessoCompareCORE.py
@@ -92,12 +92,15 @@ def main():
         parser.add_argument('--max_rows_alleles_around_cut_to_plot',  type=int, help='Maximum number of rows to report in the alleles table plot. ', default=50)
         parser.add_argument('--suppress_report',  help='Suppress output report', action='store_true')
         parser.add_argument('--place_report_in_output_folder',  help='If true, report will be written inside the CRISPResso output folder. By default, the report will be written one directory up from the report output.', action='store_true')
+        parser.add_argument('--zip_output', help="If set, the output will be placed in a zip folder.", action='store_true')
         parser.add_argument('--debug', help='Show debug messages', action='store_true')
 
         args = parser.parse_args()
         debug_flag = args.debug
 
-
+        if args.zip_output and not args.place_report_in_output_folder:
+            logger.warn('Invalid arguement combination: If zip_output is True then place_report_in_output_folder must also be True. Setting place_report_in_output_folder to True.')
+            args.place_report_in_output_folder = True
         #check that the CRISPResso output is present and fill amplicon_info
         quantification_file_1, amplicon_names_1, amplicon_info_1=CRISPRessoShared.check_output_folder(args.crispresso_output_folder_1)
         quantification_file_2, amplicon_names_2, amplicon_info_2=CRISPRessoShared.check_output_folder(args.crispresso_output_folder_2)
@@ -433,6 +436,9 @@ def get_plot_title_with_ref_name(plotTitle, refName):
 
         CRISPRessoShared.write_crispresso_info(crispresso2Compare_info_file, crispresso2_info)
 
+        if args.zip_output:
+            CRISPRessoShared.zip_results(OUTPUT_DIRECTORY)
+
         info('Analysis Complete!')
         print(CRISPRessoShared.get_crispresso_footer())
         sys.exit(0)

diff --git a/CRISPResso2/CRISPRessoPooledCORE.py b/CRISPResso2/CRISPRessoPooledCORE.py
@@ -282,11 +282,14 @@ def main():
         args = parser.parse_args()
 
         crispresso_options = CRISPRessoShared.get_crispresso_options()
-        options_to_ignore = {'fastq_r1', 'fastq_r2', 'amplicon_seq', 'amplicon_name', 'output_folder', 'name'}
+        options_to_ignore = {'fastq_r1', 'fastq_r2', 'amplicon_seq', 'amplicon_name', 'output_folder', 'name', 'zip_output'}
         crispresso_options_for_pooled = list(crispresso_options-options_to_ignore)
 
         files_to_remove = []
 
+        if args.zip_output and not args.place_report_in_output_folder:
+            logger.warn('Invalid arguement combination: If zip_output is True then place_report_in_output_folder must also be True. Setting place_report_in_output_folder to True.')
+            args.place_report_in_output_folder = True
 
         info('Checking dependencies...')
 
@@ -1595,6 +1598,9 @@ def default_sigpipe():
             crispresso2_info_file, crispresso2_info,
         )
 
+        if args.zip_output:
+            CRISPRessoShared.zip_results(OUTPUT_DIRECTORY)
+
         info('All Done!')
         print(CRISPRessoShared.get_crispresso_footer())
         sys.exit(0)

diff --git a/CRISPResso2/CRISPRessoPooledWGSCompareCORE.py b/CRISPResso2/CRISPRessoPooledWGSCompareCORE.py
@@ -161,6 +161,7 @@ def main():
             help='Show debug messages',
             action='store_true',
         )
+        parser.add_argument('--zip_output', help="If set, the output will be placed in a zip folder.", action='store_true')
 
         args = parser.parse_args()
         debug_flag = args.debug
@@ -174,6 +175,10 @@ def main():
             'debug',
         ]
 
+        if args.zip_output and not args.place_report_in_output_folder:
+            logger.warn('Invalid arguement combination: If zip_output is True then place_report_in_output_folder must also be True. Setting place_report_in_output_folder to True.')
+            args.place_report_in_output_folder = True
+
         sample_1_name = CRISPRessoShared.slugify(args.sample_1_name)
         sample_2_name = CRISPRessoShared.slugify(args.sample_2_name)
 
@@ -366,6 +371,9 @@ def main():
             crispresso2Compare_info_file, crispresso2_info,
         )
 
+        if args.zip_output:
+            CRISPRessoShared.zip_results(OUTPUT_DIRECTORY)
+
         info('All Done!')
         print(CRISPRessoShared.get_crispresso_footer())
         sys.exit(0)