Revision 14540
Added by Aaron Marcuse-Kubitza over 10 years ago
| trunk/lib/tnrs.py | ||
|---|---|---|
| 1 | 1 |
# TNRS |
| 2 | 2 |
|
| 3 |
import json |
|
| 3 | 4 |
import os.path |
| 4 | 5 |
import re |
| 5 | 6 |
import sys |
| ... | ... | |
| 131 | 132 |
+response) |
| 132 | 133 |
return match.groups() |
| 133 | 134 |
|
| 134 |
def single_tnrs_request(names, debug=False, cumulative_profiler=None): |
|
| 135 |
def single_tnrs_request(names, debug=False, cumulative_profiler=None, |
|
| 136 |
use_tnrs_export=True): |
|
| 135 | 137 |
''' |
| 136 | 138 |
Note that names containing only whitespace characters (after gwt_encode()) |
| 137 | 139 |
are ignored by TNRS and do not receive a response row. Thus, you should |
| 138 | 140 |
always match up the Name_submitted returned by TNRS with the actual |
| 139 | 141 |
submitted name to determine the corresponding TNRS response row. |
| 142 |
|
|
| 143 |
@param use_tnrs_export whether to use TNRS's TSV export feature, which |
|
| 144 |
currently returns incorrect selected matches (vegpath.org/issues/943) |
|
| 140 | 145 |
''' |
| 141 | 146 |
name_ct = len(names) |
| 142 | 147 |
assert name_ct <= max_names |
| ... | ... | |
| 200 | 205 |
# lengths depending on the taxonomic ranks present in the provided taxon |
| 201 | 206 |
# name. the extra download step is therefore necessary. |
| 202 | 207 |
|
| 203 |
debug_log('Prepare download')
|
|
| 204 |
request = download_request_template.replace('[key]', key_enc)
|
|
| 205 |
response, response_info = do_request(request) |
|
| 206 |
csv_url, = parse_response('download', download_response_pattern,
|
|
| 207 |
response, response, response_info) |
|
| 208 |
csv_url += download_url_suffix |
|
| 209 |
debug_log('csv_url', csv_url)
|
|
| 210 |
|
|
| 211 |
debug_log('Download')
|
|
| 212 |
request_obj = urllib2.Request(csv_url) |
|
| 213 |
debug_log('request URL', str(csv_url))
|
|
| 214 |
debug_log('request info', str(request_obj.header_items()))
|
|
| 215 |
debug_log('request str', str(request_obj.get_data()))
|
|
| 216 |
response = urllib2.urlopen(request_obj) |
|
| 217 |
response_info = str(response.info()) |
|
| 218 |
debug_log('response info', response_info)
|
|
| 219 |
return TnrsOutputStream(response) |
|
| 208 |
if use_tnrs_export: |
|
| 209 |
debug_log('Prepare download')
|
|
| 210 |
request = download_request_template.replace('[key]', key_enc)
|
|
| 211 |
response, response_info = do_request(request) |
|
| 212 |
csv_url, = parse_response('download', download_response_pattern,
|
|
| 213 |
response, response, response_info) |
|
| 214 |
csv_url += download_url_suffix |
|
| 215 |
debug_log('csv_url', csv_url)
|
|
| 216 |
|
|
| 217 |
debug_log('Download')
|
|
| 218 |
request_obj = urllib2.Request(csv_url) |
|
| 219 |
debug_log('request URL', str(csv_url))
|
|
| 220 |
debug_log('request info', str(request_obj.header_items()))
|
|
| 221 |
debug_log('request str', str(request_obj.get_data()))
|
|
| 222 |
response = urllib2.urlopen(request_obj) |
|
| 223 |
response_info = str(response.info()) |
|
| 224 |
debug_log('response info', response_info)
|
|
| 225 |
return TnrsOutputStream(response) |
|
| 226 |
else: |
|
| 227 |
debug_log('Expand')
|
|
| 228 |
request = ('7|0|6|http://tnrs.iplantcollaborative.org/tnrsdemo/|\
|
|
| 229 |
1E87C78041CEFBF0992F46BDF84D7D60|org.iplantc.tnrs.demo.client.SearchService|\ |
|
| 230 |
requestGroupMembers|java.lang.String/2004016611|{"group":"0", \
|
|
| 231 |
"email":"tnrs@lka5jjs.orv", "key":"[key]", "source_sorting":"false", \ |
|
| 232 |
"taxonomic_constraint":"false"}|1|2|3|4|1|5|6|'.replace('[key]', key_enc))
|
|
| 233 |
response, response_info = do_request(request) |
|
| 234 |
data_gwt, = parse_response('expand', r'^//OK\[1,\["(.*)"\],0,7\]$',
|
|
| 235 |
response, response, response_info) |
|
| 236 |
data_json = json.loads(gwt_decode(data_gwt)) |
|
| 237 |
raise NotImplementedError() |
|
| 220 | 238 |
finally: |
| 221 | 239 |
profiler.stop(name_ct) |
| 222 | 240 |
sys.stderr.write(profiler.msg()+'\n') |
Also available in: Unified diff
lib/tnrs.py: added option to avoid using TNRS's TSV export feature, which currently returns incorrect selected matches (vegpath.org/issues/943). this has been implemented up through the GWT/JSON decoding.