| Index: trunk/tools/wsor/message_templates/run_postings_and_metrics.py |
| — | — | @@ -30,52 +30,59 @@ |
| 31 | 31 | Execution body of main |
| 32 | 32 | """ |
| 33 | 33 | def main(args): |
| | 34 | + |
| | 35 | + # Set flags for template indices to process |
| | 36 | + template_indices = {1 : False, 2 : False, 3 : False, |
| | 37 | + 60 : False, 61 : False, 62 : False, 63 : False, 64 : False, 65 : False, 66 : False, 67 : False, 68 : False, 69 : False, |
| | 38 | + 70 : False, 71 : False, 72 : False, 73 : False, 74 : False, 75 : False, 76 : False, 77 : False, |
| | 39 | + 84 : False, 85 : False, 86 : False, 99 : False, 100 : False, 101 : False, 102 : False, 103 : False, 104 : False, |
| | 40 | + 105 : False, 106 : False, 107 : False, 108 : False, 109 : False, 110 : False, 111 : False, 112 : False, 113 : False, |
| | 41 | + 114 : False, 115 : False, 116 : False, 78 : False, 79 : False, 81 : False, 82 : False, 87 : False, 88 : False, |
| | 42 | + 89 : True, 90 : True, 91 : True, 92 : True, 93 : True, 94 : True, 95 : True, 96 : True, 97 : True, 98 : True |
| | 43 | + } |
| 34 | 44 | |
| 35 | | - test_keys = ['Huggle_3_z60', 'Huggle_3_z61', 'Huggle_3_z62', 'Huggle_3_z63', 'Huggle_3_z64', 'Huggle_3_z65', 'Huggle_3_z66', \ |
| 36 | | - 'Huggle_3_z67', 'Huggle_3_z68', 'Huggle_3_z69', 'Huggle_3_z70', 'Huggle_3_z71', 'Huggle_3_z72', 'Huggle_3_z73', \ |
| 37 | | - 'Huggle_3_z74', 'Huggle_3_z75', 'Huggle_3_z76', \ |
| 38 | | - 'Huggle_3_z77', 'Huggle_3_z1', 'Huggle_3_z2', 'Huggle_3_z3'] # "Huggle 1 Portugese" |
| 39 | | - |
| 40 | | - gather_data = {'Huggle_3_z60' : False, 'Huggle_3_z61' : False, 'Huggle_3_z62' : False, 'Huggle_3_z63' : False, 'Huggle_3_z64' : False, 'Huggle_3_z65' : False, 'Huggle_3_z66' : False, 'Huggle_3_z67' : False, \ |
| 41 | | - 'Huggle_3_z68' : False, 'Huggle_3_z69' : False, 'Huggle_3_z70' : False, 'Huggle_3_z71' : False, 'Huggle_3_z72' : False, 'Huggle_3_z73' : False, 'Huggle_3_z74' : False, 'Huggle_3_z75' : False, \ |
| 42 | | - 'Huggle_3_z76' : False, 'Huggle_3_z77' : False, \ |
| 43 | | - 'Huggle_3_z1' : True, 'Huggle_3_z2' : True, 'Huggle_3_z3' : True} |
| 44 | | - |
| 45 | | - templates = {'Huggle_3_z60' : 'z60', 'Huggle_3_z61' : 'z61', 'Huggle_3_z62' : 'z62', 'Huggle_3_z63' : 'z63', 'Huggle_3_z64' : 'z64', 'Huggle_3_z65' : 'z65', 'Huggle_3_z66' : 'z66', 'Huggle_3_z67' : 'z67', \ |
| 46 | | - 'Huggle_3_z68' : 'z68', 'Huggle_3_z69' : 'z69', 'Huggle_3_z70' : 'z70', 'Huggle_3_z71' : 'z71', 'Huggle_3_z72' : 'z72', 'Huggle_3_z73' : 'z73', 'Huggle_3_z74' : 'z74', 'Huggle_3_z75' : 'z75', \ |
| 47 | | - 'Huggle_3_z76' : 'z76', 'Huggle_3_z77' : 'z77', \ |
| 48 | | - 'Huggle_3_z1' : 'z1', 'Huggle_3_z2' : 'z2', 'Huggle_3_z3' : 'z3'} |
| 49 | | - |
| 50 | | - start_times = {'Huggle_3_z60' : '20111018000000', 'Huggle_3_z61' : '20111018000000', 'Huggle_3_z62' : '20111018000000', 'Huggle_3_z63' : '20111018000000', 'Huggle_3_z64' : '20111018000000', 'Huggle_3_z65' : '20111018000000', \ |
| 51 | | - 'Huggle_3_z66' : '20111018000000', 'Huggle_3_z67' : '20111018000000', 'Huggle_3_z68' : '20111018000000', 'Huggle_3_z69' : '20111018000000', 'Huggle_3_z70' : '20111018000000', 'Huggle_3_z71' : '20111018000000', \ |
| 52 | | - 'Huggle_3_z72' : '20111018000000', 'Huggle_3_z73' : '20111018000000', 'Huggle_3_z74' : '20111018000000', 'Huggle_3_z75' : '20111018000000', 'Huggle_3_z76' : '20111018000000', 'Huggle_3_z77' : '20111018000000', \ |
| 53 | | - 'Huggle_3_z1' : '20111027000000', 'Huggle_3_z2' : '20111027000000', 'Huggle_3_z3' : '20111027000000'} |
| 54 | | - |
| 55 | | - end_times = {'Huggle_3_z60' : '20111119000000', 'Huggle_3_z61' : '20111119000000', 'Huggle_3_z62' : '20111119000000', 'Huggle_3_z63' : '20111119000000', 'Huggle_3_z64' : '20111119000000', 'Huggle_3_z65' : '20111119000000', \ |
| 56 | | - 'Huggle_3_z66' : '20111119000000', 'Huggle_3_z67' : '20111119000000', 'Huggle_3_z68' : '20111119000000', 'Huggle_3_z69' : '20111119000000', 'Huggle_3_z70' : '20111119000000', 'Huggle_3_z71' : '20111119000000', \ |
| 57 | | - 'Huggle_3_z72' : '20111119000000', 'Huggle_3_z73' : '20111119000000', 'Huggle_3_z74' : '20111119000000', 'Huggle_3_z75' : '20111119000000', 'Huggle_3_z76' : '20111119000000', 'Huggle_3_z77' : '20111119000000', |
| 58 | | - 'Huggle_3_z1' : '20111128000000', 'Huggle_3_z2' : '20111128000000', 'Huggle_3_z3' : '20111128000000'} |
| 59 | | - |
| 60 | 45 | # Run postings and metrics |
| 61 | 46 | |
| 62 | | - generator = 'editcounts' |
| | 47 | + generator = 'editcount' |
| 63 | 48 | postings_cmd = './postings -h db1047 --start=%(start_time)s --end=%(end_time)s --comment="\(\[\[WP:HG\|HG\]\]\)" --message="{{%(template)s}}" --outfilename postings_%(file_name)s.tsv' |
| 64 | | - metrics_cmd = 'cat ./output/postings_%(file_name)s.tsv | ./metrics -h db1047 --header --outfilename metrics_%(file_name)s.tsv %(generator)s' |
| | 49 | + metrics_cmd = 'cat ./output/postings_%(file_name)s.tsv | ./metrics -h db1047 --header --outfilename metrics_%(file_name)s_%(fname_generator)s.tsv %(generator)s' |
| 65 | 50 | |
| 66 | | - for key in test_keys: |
| 67 | | - |
| 68 | | - if gather_data[key]: |
| 69 | | - logging.info('Generating postings for %s' % key) |
| 70 | | - filename_part = start_times[key][4:8] + '_' + end_times[key][4:8] + '_' + templates[key] |
| | 51 | + for key in template_indices: |
| | 52 | + |
| | 53 | + name, start_ts, end_ts = get_experiment(key) |
| | 54 | + template_name = 'z' + str(key) |
| | 55 | + |
| | 56 | + if template_indices[key]: |
| | 57 | + |
| | 58 | + logging.info('Generating postings for %s' % template_name) |
| | 59 | + filename_part = start_ts[4:8] + '_' + end_ts[4:8] + '_' + template_name |
| 71 | 60 | |
| 72 | | - # os.system(postings_cmd % {'start_time' : start_times[key], 'end_time' : end_times[key], 'template' : templates[key], 'file_name' : filename_part}) |
| 73 | | - os.system(metrics_cmd % {'file_name' : filename_part, 'file_name' : filename_part, 'generator' : generator}) |
| | 61 | + os.system(postings_cmd % {'start_time' : start_ts, 'end_time' : end_ts, 'template' : template_name, 'file_name' : filename_part}) |
| | 62 | + # os.system(metrics_cmd % {'file_name' : filename_part, 'file_name' : filename_part, 'generator' : generator, 'fname_generator' : generator}) |
| 74 | 63 | else: |
| 75 | | - logging.info('Skipping postings for %s' % key) |
| | 64 | + logging.info('Skipping postings for %s' % template_name) |
| 76 | 65 | |
| 77 | 66 | return 0 |
| 78 | 67 | |
| 79 | 68 | """ |
| | 69 | + Returns the experiment name and start and end timestamps corresponding to the key |
| | 70 | +""" |
| | 71 | +def get_experiment(index): |
| | 72 | + |
| | 73 | + if index >= 60 and index <= 77: |
| | 74 | + return 'Huggle_3', '20111018000000', '20111119000000' |
| | 75 | + elif index >= 1 and index <= 3: |
| | 76 | + return 'Huggle_1_Portuguese', '20111027000000', '20111128000000' |
| | 77 | + elif (index >= 84 and index <= 86) or (index >= 99 and index <= 106): |
| | 78 | + return 'Huggle_Short_1_and_2', '20111108000000', '20111202000000' |
| | 79 | + elif index >= 107 and index <= 116: |
| | 80 | + return 'Huggle_Short_2', '20111122000000', '20111222000000' |
| | 81 | + elif (index >= 77 and index <= 79) or (index >= 81 and index <= 82): |
| | 82 | + return 'Twinkle_1', '20111109000000', '20111209000000' |
| | 83 | + elif index >= 87 and index <= 98: |
| | 84 | + return 'XLinkBot', '20111117000000', '20111217000000' |
| | 85 | + |
| | 86 | +""" |
| 80 | 87 | Call main, exit when execution is complete |
| 81 | 88 | |
| 82 | 89 | """ |