comparison generate-foo.sh @ 9:e18096659b38 draft

"planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/openms commit ddf41e8bda1ba065f5cdec98e93dee8165ffc1b9"
author galaxyp
date Thu, 03 Sep 2020 16:19:26 +0000
parents
children 7951e3db2efd
comparison
equal deleted inserted replaced
8:66d8e5a850f1 9:e18096659b38
1 #!/usr/bin/env bash
2
3 # parse test definitions from OpenMS sources for a tool with a given id
4 function get_tests2 {
5 id=$1
6 >&2 echo "generate tests for $id"
7 echo '<xml name="autotest_'"$id"'">'
8
9 # get the tests from the CMakeLists.txt
10 # 1st remove some tests
11 # - Filefilter with empty select_palarity value (empty is not in the list of allowed options)
12 # - MassTraceExtractor with outdated ini file leading to wrong parameters https://github.com/OpenMS/OpenMS/issues/4386
13 # - OpenSwathMzMLFileCacher with -convert_back argumen https://github.com/OpenMS/OpenMS/issues/4399
14 # - IDRipper PATH gets empty causing problems. TODO But overall the option needs to be handled differentlt
15 # - several tools with duplicated input (leads to conflict when linking)
16 # - TOFCalibration inputs we extension (also in prepare_test_data) https://github.com/OpenMS/OpenMS/pull/4525
17 # - MaRaCluster with -consensus_out (parameter blacklister: https://github.com/OpenMS/OpenMS/issues/4456)
18 # - FileMerger with mixed dta dta2d input (ftype can not be specified in the test, dta can not be sniffed)
19 # - some input files are originally in a subdir (degenerated cases/), but not in test-data
20 # - SeedListGenerator: https://github.com/OpenMS/OpenMS/issues/4404
21 # - OpenSwathAnalyzer 9/10: cachedMzML (not supported yet)
22 CMAKE=$(cat $OPENMSGIT/src/tests/topp/CMakeLists.txt $OPENMSGIT/src/tests/topp/THIRDPARTY/third_party_tests.cmake |
23 sed 's@${DATA_DIR_SHARE}/@@g' |
24 grep -v 'OpenSwathMzMLFileCacher .*-convert_back' |
25 sed 's/${TMP_RIP_PATH}/""/' |
26 sed 's@TOFCalibration_ref_masses @TOFCalibration_ref_masses.txt @g; s@TOFCalibration_const @TOFCalibration_const.csv @' |
27 grep -v "MaRaClusterAdapter.*-consensus_out"|
28 grep -v "FileMerger_1_input1.dta2d.*FileMerger_1_input2.dta " |
29 sed 's@degenerate_cases/@@g' |
30 grep -v 'TOPP_SeedListGenerator_3"' |
31 egrep -v 'TOPP_OpenSwathAnalyzer_test_3"|TOPP_OpenSwathAnalyzer_test_4"')
32
33
34 # grep -v 'FileFilter.*-spectra:select_polarity ""' |
35 # grep -v 'MassTraceExtractor_2.ini ' |
36 # grep -v "FileMerger_6_input2.mzML.*FileMerger_6_input2.mzML" |
37 # grep -v "IDMerger_1_input1.idXML.*IDMerger_1_input1.idXML" |
38 # grep -v "degenerated_empty.idXML.*degenerated_empty.idXML" |
39 # grep -v "FeatureLinkerUnlabeledKD_1_output.consensusXML.*FeatureLinkerUnlabeledKD_1_output.consensusXML" |
40 # grep -v "FeatureLinkerUnlabeledQT_1_output.consensusXML.*FeatureLinkerUnlabeledQT_1_output.consensusXML" |
41
42 # 1st part is a dirty hack to join lines containing a single function call, e.g.
43 # addtest(....
44 # ....)
45 echo "$CMAKE" | sed 's/#.*//; s/^\s*//; s/\s*$//' | grep -v "^#" | grep -v "^$" | awk '{printf("%s@NEWLINE@", $0)}' | sed 's/)@NEWLINE@/)\n/g' | sed 's/@NEWLINE@/ /g' |
46 grep -iE "add_test\(\"(TOPP|UTILS)_.*/$id " | egrep -v "_prepare\"|_convert|WRITEINI|WRITECTD|INVALIDVALUE" | while read -r line
47 do
48 line=$(echo "$line" | sed 's/add_test("\([^"]\+\)"/\1/; s/)$//; s/\${TOPP_BIN_PATH}\///g;s/\${DATA_DIR_TOPP}\///g; s#THIRDPARTY/##g')
49 # >&2 echo $line
50 test_id=$(echo "$line" | cut -d" " -f 1)
51 tool_id=$(echo "$line" | cut -d" " -f 2)
52 if [[ $test_id =~ _out_?[0-9]? ]]; then
53 >&2 echo " skip $test_id $line"
54 continue
55 fi
56 if [[ ${id,,} != ${tool_id,,} ]]; then
57 >&2 echo " skip $test_id ($id != $tool_id) $line"
58 continue
59 fi
60
61 #remove tests with set_tests_properties(....PROPERTIES WILL_FAIL 1)
62 if grep -lq "$test_id"'\".* PROPERTIES WILL_FAIL 1' $OPENMSGIT/src/tests/topp/CMakeLists.txt $OPENMSGIT/src/tests/topp/THIRDPARTY/third_party_tests.cmake; then
63 >&2 echo " skip failing "$test_id
64 continue
65 fi
66 tes=" <test>\n"
67 line=$(fix_tmp_files "$line")
68 line=$(unique_files "$line")
69 # >&2 echo $line
70 #if there is an ini file then we use this to generate the test
71 #otherwise the ctd file is used
72 #other command line parameters are inserted later into this xml
73 if grep -lq "\-ini" <<<"$line"; then
74 ini=$(echo $line | sed 's/.*-ini \([^ ]\+\).*/\1/')
75 ini="test-data/$ini"
76 else
77 ini="ctd/$tool_id.ctd"
78 fi
79 cli=$(echo $line |cut -d" " -f3- | sed 's/-ini [^ ]\+//')
80
81 ctdtmp=$(mktemp)
82 #echo python3 fill_ctd_clargs.py --ctd $ini $cli
83 # using eval: otherwise for some reason quoted values are not used properly ('A B' -> ["'A", "B'"])
84 # >&2 echo "python3 fill_ctd_clargs.py --ctd $ini $cli"
85 eval "python3 fill_ctd_clargs.py --ctd $ini $cli" > "$ctdtmp"
86 # echo $ctdtmp
87 # >&2 cat $ctdtmp
88 testtmp=$(mktemp)
89 python3 $CTDCONVERTER/convert.py galaxy -i $ctdtmp -o $testtmp -s tools_blacklist.txt -f "$FILETYPES" -m macros.xml -t tool.conf -p hardcoded_params.json --tool-version $VERSION --test-only --test-unsniffable csv tsv txt dta dta2d edta mrm splib > /dev/null
90 cat $testtmp | grep -v '<output.*file=""' # | grep -v 'CHEMISTRY/'
91 rm $ctdtmp $testtmp
92
93 #> /dev/null
94
95 #rm $testtmp
96 done
97 echo '</xml>'
98 }
99
100 #some tests use the same file twice which does not work in planemo tests
101 #hence we create symlinks for each file used twice
102 function unique_files {
103 line=$@
104 for arg in $@
105 do
106 if [[ ! -f "test-data/$arg" ]]; then
107 continue
108 fi
109 cnt=$(grep -c $arg <<< $(echo "$line" | tr ' ' '\n'))
110 while [[ $cnt -gt 1 ]]; do
111 new_arg=$(echo $arg | sed "s/\(.*\)\./\1_$cnt./")
112 ln -fs $arg test-data/$new_arg
113 line=$(echo $line | sed "s/\($arg.*\)$arg/\1$new_arg/")
114 cnt=$(grep -c $arg <<< $(echo "$line" | tr ' ' '\n'))
115 done
116 done
117
118 echo $line
119 }
120
121 # options of out_type selects need to be fixed to Galaxy data types
122 function fix_out_type {
123 grep "^$1" "$2" | awk '{print $2}'
124 }
125
126 #OpenMS tests output to tmp files and compare with FuzzyDiff to the expected file.
127 #problem: the extension of the tmp files is unusable for test generation.
128 #unfortunately the extensions used in the DIFF lines are not always usable for the CLI
129 #(e.g. for prepare_test_data, e.g. CLI expects csv but test file is txt)
130 #this function replaces the tmp file by the expected file.
131 function fix_tmp_files {
132 # >&2 echo "FIX $line"
133 ret=""
134 for a in $@; do
135 if [[ ! $a =~ .tmp$ ]]; then
136 ret="$ret $a"
137 continue
138 fi
139 # >&2 echo " a "$a
140 g=$(cat $OPENMSGIT/src/tests/topp/CMakeLists.txt $OPENMSGIT/src/tests/topp/THIRDPARTY/third_party_tests.cmake | awk '{printf("%s@NEWLINE@", $0)}' | sed 's/)@NEWLINE@/)\n/g' | sed 's/@NEWLINE@/ /g' | grep '\${DIFF}.*'"$a")
141 # >&2 echo " g "$g
142 in1=$(sed 's/.*-in1 \([^ ]\+\).*/\1/' <<<$g)
143 # >&2 echo " in1 "$in1
144 if [[ "$a" != "$in1" ]]; then
145 ret="$ret $a"
146 continue
147 fi
148 in2=$(sed 's/.*-in2 \([^ ]\+\).*/\1/' <<<$g)
149 in2=$(basename $in2 | sed 's/)$//')
150 # >&2 echo " in2 "$in2
151 if [[ -f "test-data/$in2" ]]; then
152 ln -fs "$in1" "test-data/$in2"
153 ret="$ret $in2"
154 else
155 ret="$ret $a"
156 fi
157 done
158 # >&2 echo "--> $ret"
159 echo "$ret"
160 }
161
162 function link_tmp_files {
163 # note this also considers commented lines (starting with a #)
164 # because of tests where the diff command is commented and we
165 # still want to use the extension of these files
166 cat $OPENMSGIT/src/tests/topp/CMakeLists.txt $OPENMSGIT/src/tests/topp/THIRDPARTY/third_party_tests.cmake | sed 's/^\s*//; s/\s*$//' | grep -v "^$" | awk '{printf("%s@NEWLINE@", $0)}' | sed 's/)@NEWLINE@/)\n/g' | sed 's/@NEWLINE@/ /g' | grep "\${DIFF}" | while read -r line
167 do
168 in1=$(sed 's/.*-in1 \([^ ]\+\).*/\1/' <<<$line)
169 in1=$(basename $in1 | sed 's/)$//')
170 in2=$(sed 's/.*-in2 \([^ ]\+\).*/\1/' <<<$line)
171 in2=$(basename $in2 | sed 's/)$//')
172 if [[ "$in1" == "$in2" ]]; then
173 >&2 echo "not linking equal $in1 $in2"
174 continue
175 fi
176 ln -f -s $in1 test-data/$in2
177 done
178 for i in test-data/*.tmp
179 do
180 if [ ! -e test-data/$(basename $i .tmp) ]; then
181 ln -s $(basename $i) test-data/$(basename $i .tmp)
182 #ln -s $(basename $i) test-data/$(basename $i .tmp)
183 else
184 ln -fs $(basename $i) test-data/$(basename $i .tmp)
185 fi
186 done
187 }
188
189
190
191 # parse data preparation calls from OpenMS sources for a tool with a given id
192 function prepare_test_data {
193 # id=$1
194 # | egrep -i "$id\_.*[0-9]+(_prepare\"|_convert)?"
195 cat $OPENMSGIT/src/tests/topp/CMakeLists.txt $OPENMSGIT/src/tests/topp/THIRDPARTY/third_party_tests.cmake | sed 's/#.*$//'| sed 's/^\s*//; s/\s*$//' | grep -v "^$" | awk '{printf("%s@NEWLINE@", $0)}' | sed 's/)@NEWLINE@/)\n/g' | sed 's/@NEWLINE@/ /g' |
196 sed 's/degenerate_cases\///' |
197 egrep -v "WRITEINI|WRITECTD|INVALIDVALUE|DIFF" |
198 grep add_test |
199 egrep "TOPP|UTILS" |
200 sed 's@${DATA_DIR_SHARE}/@@g;'|
201 sed 's@${TMP_RIP_PATH}@dummy2.tmp@g'|
202 sed 's@TOFCalibration_ref_masses @TOFCalibration_ref_masses.txt @g; s@TOFCalibration_const @TOFCalibration_const.csv @'|
203 while read line
204 do
205 test_id=$(echo "$line" | sed 's/add_test(//; s/"//g; s/)[^)]*$//; s/\${TOPP_BIN_PATH}\///g;s/\${DATA_DIR_TOPP}\///g; s#THIRDPARTY/##g' | cut -d" " -f1)
206
207 if grep -lq "$test_id"'\".* PROPERTIES WILL_FAIL 1' $OPENMSGIT/src/tests/topp/CMakeLists.txt $OPENMSGIT/src/tests/topp/THIRDPARTY/third_party_tests.cmake; then
208 >&2 echo " skip failing "$test_id
209 continue
210 fi
211
212 line=$(echo "$line" | sed 's/add_test("//; s/)[^)]*$//; s/\${TOPP_BIN_PATH}\///g;s/\${DATA_DIR_TOPP}\///g; s#THIRDPARTY/##g' | cut -d" " -f2-)
213 # line="$(fix_tmp_files $line)"
214 echo "$line > $test_id.stdout 2> $test_id.stderr"
215 echo "if [[ \"\$?\" -ne \"0\" ]]; then >&2 echo '$test_id failed'; >&2 echo -e \"stderr:\n\$(cat $test_id.stderr | sed 's/^/ /')\"; echo -e \"stdout:\n\$(cat $test_id.stdout)\";fi"
216 done
217 }