MARC command line (marcli) is a tool to parse MARC files from the command line.
The goal of marcli is to allow a user to parse large MARC files in either MARC binary or MARC XML form and provide basic functionality to find specific records within a file.
On the Mac or Linux the easiest way to install marcli is via Homebrew (thank you Eric Phetteplace):
brew install marcli
marcli -file yourfile.mrc
Or by downloading the binary for your OS from the releases tab and marking the downloaded file as an executable:
curl -LO https://github.com/hectorcorrea/marcli/releases/latest/download/marcli
chmod u+x marcli
./marcli -file yourfile.mrc
Notice that if you install marcli via Homebrew the executable will be on your path and you can run it just by typing marcli from any folder, whereas if you install it via cURL you need to indicate the path ./marcli.
Output MARC data to the console in a line delimited format (marcli automatically detects whether the file provided is in MARC XML or MARC binary):
./marcli -file data/test_1a.mrc
./marcli -file data/test_10.xml
You can use the -match parameter to get only the records that match a given string, for example the code below extracts MARC records that contain the string "wildlife"
./marcli -file data/test_10.mrc -match wildlife
Extracts MARC records on file that contain the string "wildlife" but outputs only fields "LDR,001,040,245a,650" for each record, LDR means the leader of the MARC record. In the -fields parameter a letter (or letters) after the field tag indicates to output only those subfields. For example "907xz" means output subfield "x" and "z" in field "907".
./marcli -file data/test_10.mrc -match wildlife -fields LDR,010,040,245a,650
The -matchRegEx parameter can be used to pass a regular expression instead of a value to select the records that will be matched, for example the following will print only those records that have values that look like dates for March 2006 (03-\d\d-06):
./marcli -file data/test_10.xml -matchRegEx '.*03-\d\d-06.*'
The -matchFields parameter can be used to limit the fields where the match will be made:
./marcli -file data/test_10.mrc -match web -matchFields 530
You can also use the exclude option to indicate fields to exclude from the output (notice that only full fields are supported here, e.g. 970 is accepted but not 970a)
You can also filter based on the presence of certain fields in the MARC record (regardless of their value), for example the following will only output records that have a MARC 110 field:
./marcli -file data/test_10.mrc -hasFields 110
By default the output is in Mnemonic MARC (.mrk), which is a human readable format. You can use the format parameter to output MARC XML, MARC JSON, or MARC binary instead. Notice that not all the features are available in all the formats.
You can use count-only as the format if you only want a count of the number of records on the file. If you use the match parameter it will report only the number of records that match the criteria.
You can also pass start and count parameters to output only a range of MARC records.
Files under ./data/ are small MARC files that I use for testing.
- test_10.mrc has 10 MARC records (MARC binary)
- test_1a.mrc is the first record of test_10.mrc (MARC binary)
- test_1b.mrc is the second record of test_10.mrc (MARC binary)
- test_10.xml same as test_10.mrc but in MARC XML.
Download the code and play with it:
cd ~/src
git clone https://github.com/hectorcorrea/marcli.git
cd marcli/cmd/marcli
go build
./marcli -file ~/src/marcli/data/test_1a.mrc
cmd/marclicontains the code for the command line interface.pkg/marccontains the code to parse MARC files.
If you find an issue parsing MARC files with marcli feel free to submit an issue with details of the error, and if possible a sample file or contact me by email at hector@hectorcorrea.com
- Code4Lib 2021 lightning talk (slides and video)
- Understanding MARC: https://www.loc.gov/marc/umb/