diff options
Diffstat (limited to 'skate/cmd/skate-reduce')
| -rw-r--r-- | skate/cmd/skate-reduce/main.go | 93 |
1 files changed, 49 insertions, 44 deletions
diff --git a/skate/cmd/skate-reduce/main.go b/skate/cmd/skate-reduce/main.go index 098240a..fbb4f57 100644 --- a/skate/cmd/skate-reduce/main.go +++ b/skate/cmd/skate-reduce/main.go @@ -5,50 +5,55 @@ // Support various "modes", e.g. exact, verify, ref, bref, wiki. Each mode may // work on one or two files, and may need extra args. // -// * exact | takes two (key, doc) TSV files (one for releases, one for refs) and -// | will emit biblioref docs relating *one* element from releases with *all* -// | elements from ref; this is for "doi", "pmid" and other id matches, where no -// | further checks are necessary. The match reason, e.g. "doi" needs to be -// | supplied. -// | -// | $ skate-reduce -m exact -r doi -F a.tsv -L b.tsv -// | -// | -// * fuzzy | takes two (key, doc) TSV files (one for release, one for refs), -// | runs verification within a group and will emit biblioref. -// | -// | $ skate-reduce -m fuzzy -F a.tsv -L b.tsv -// | -// | -// * ref | takes a single file with clusters containing releases and refs and -// | will emit verification results (deprecated). -// | -// | $ skate-reduce -m ref < a.ndj -// | -// | -// * bref | same as ref, but generate a biblioref file as output (deprecated). -// | -// | $ skate-reduce -m bref < a.ndj -// | -// | -// * wiki | zippy mode for releases and wikipedia inputs. -// | -// | $ skate-reduce -m wiki -L a.ndj -W b.ndj -// | -// | -// * oledt | zippy mode for releases and OL inputs, dumps table for debugging. -// | -// | $ skate-reduce -m oled -F a.ndj -O b.ndj -// | -// | -// * oled | zippy mode for releases and OL inputs, emit biblioref. -// | -// | $ skate-reduce -m oled -F a.ndj -O b.ndj -// | -// | -// * rere | zippy mode for OL release and ref (as release) inputs, emit biblioref. -// | -// | $ skate-reduce -m rere -O a.ndj -F b.ndj +// * exact | takes two (key, doc) TSV files (one for releases, one for refs) and +// | will emit biblioref docs relating *one* element from releases with *all* +// | elements from ref; this is for "doi", "pmid" and other id matches, where no +// | further checks are necessary. The match reason, e.g. "doi" needs to be +// | supplied. +// | +// | $ skate-reduce -m exact -r doi -F a.tsv -L b.tsv +// | +// | +// * fuzzy | takes two (key, doc) TSV files (one for release, one for refs), +// | runs verification within a group and will emit biblioref. +// | +// | $ skate-reduce -m fuzzy -F a.tsv -L b.tsv +// | +// | +// * ref | takes a single file with clusters containing releases and refs and +// | will emit verification results (deprecated). +// | +// | $ skate-reduce -m ref < a.ndj +// | +// | +// * bref | same as ref, but generate a biblioref file as output (deprecated). +// | +// | $ skate-reduce -m bref < a.ndj +// | +// | +// * wiki | zippy mode for releases and wikipedia inputs. +// | +// | $ skate-reduce -m wiki -L a.ndj -W b.ndj +// | +// | +// * oledt | zippy mode for releases and OL inputs, dumps table for debugging. +// | +// | $ skate-reduce -m oled -F a.ndj -O b.ndj +// | +// | +// * oled | zippy mode for releases and OL inputs, emit biblioref. +// | +// | $ skate-reduce -m oled -F a.ndj -O b.ndj +// | +// | +// * rere | zippy mode for OL release and ref (as release) inputs, emit biblioref. +// | +// | $ skate-reduce -m rere -O a.ndj -F b.ndj +// | +// | +// * unmatched | join matched and unmatched reference data; do deduplicate on the fly +// | +// | $ skate-reduce -m unmatched -M a.ndj -R b // package main |