Let’s say you have a JSON file and in it is an array of values. Each value has a few properties you need to pick from to ultimately create a file name. Here is an example using bash, jq, in2csv and xargs that can be strung together in one line to create a file based on each concatenated filename.<\/p>\n\n\n\n
cat some.json \\\n | jq -r '.statements | map([.account.number,.endDate,.id] | join(\"_\") + \".pdf\")' \\\n | in2csv -f json \\\n | sed -n '1d;p' \\\n | xargs -I{} cp fubar.pdf {};<\/code><\/pre>\n\n\n\nLet’s break this down into it’s individual pieces.<\/p>\n\n\n\n
Sample JSON Input<\/h2>\n\n\n\n{\n \"statements\" : [ {\n \"id\" : \"60090438-83cc-433f-bbf9-8086f4da846f\",\n \"account\" : {\n \"number\" : \"801284289140212\"\n },\n \"endDate\" : \"2021-06-30\",\n ...\n }\n...\n]\n}<\/code><\/pre>\n\n\n\nUsing JQ To Select A Few Fields to Concatenate for a File Name<\/h2>\n\n\n\n
The following line will select three properties\/fields out of the JSON<\/p>\n\n\n\n
jq -r '.statements | map([.account.number,.endDate,.id] | join(\"_\") + \".pdf\")'<\/code><\/pre>\n\n\n\nIt will join<\/code> them by an underscore and then concatenate a .pdf<\/code> to the end of the file name.<\/p>\n\n\n\nConvert Quoted JQ JSON into CSV via In2Csv<\/h2>\n\n\n\n
The jq<\/code> command will provide quoted output. Since, there is only one value per item in the array due to the concatenation, I just want to strip out the JSON quoting.<\/p>\n\n\n\nThat is done with the following:<\/p>\n\n\n\n
in2csv -f json<\/code><\/pre>\n\n\n\nNote the use of the -f json<\/code> since this is a piped command. That will yield this warning, which we can squash next<\/p>\n\n\n\n\/usr\/local\/Cellar\/csvkit\/1.0.5_1\/libexec\/lib\/python3.8\/site-packages\/agate\/utils.py:276: UnnamedColumnWarning: Column 0 has no name. Using \"a\". <\/code><\/pre>\n\n\n\nRemove The First Line From the Bash Output<\/h2>\n\n\n\n
This piped command remove the header labeled a<\/code> by default:<\/p>\n\n\n\nsed -n '1d;p'<\/code><\/pre>\n\n\n\nCreating a Dummy Placeholder File Per File Name Generated with Xargs and Cp<\/h2>\n\n\n\n
Lastly, I want to generate a placeholder file (touch<\/code> could be used too) per file name. This command achieves that with xargs<\/code><\/p>\n\n\n\nxargs -I{} cp fubar.pdf {};<\/code><\/pre>\n\n\n\nSample output:<\/p>\n\n\n\n
\n -rw-r--r--@ 1 kevincoy staff 1676399 Jun 29 16:48 182362260357264_2021-03-31_81f1465b-b2d6-43d9-87bd-d6ede8188188.pdf\n -rw-r--r--@ 1 kevincoy staff 1676399 Jun 29 16:48 468913841109837_2021-03-31_043af01f-bd27-42a7-b37d-c7c4b7d1275f.pdf\n -rw-r--r--@ 1 kevincoy staff 1676399 Jun 29 16:48 558664362096984_2021-03-31_287c8418-98bd-4088-9645-b897c40d0ed4.pdf\n -rw-r--r--@ 1 kevincoy staff 1676399 Jun 29 16:48 881961481598397_2021-03-31_14e1e791-3e2d-4275-b27a-bf49cc2999e8.pdf\n -rw-r--r--@ 1 kevincoy staff 1676399 Jun 29 16:48 201869782345979_2021-03-31_a623f37a-dfe7-4748-af2e-6b45dbeb869b.pdf\n -rw-r--r--@ 1 kevincoy staff 1676399 Jun 29 16:48 177636620098765_2021-03-31_4007c4c2-7263-4d58-8a8f-7b5bd6aef802.pdf\n -rw-r--r--@ 1 kevincoy staff 1676399 Jun 29 16:48 812555146927128_2021-03-31_f3a06bd9-efe3-4da1-bd29-6ee93473d920.pdf <\/code><\/pre>\n\n\n\nResources: <\/p>\n\n\n\n
- https:\/\/jqplay.org\/s\/N6TboUkELM<\/li>
- https:\/\/csvkit.readthedocs.io\/en\/1.0.2\/scripts\/in2csv.html<\/li>
- https:\/\/stackoverflow.com\/questions\/604864\/print-a-file-skipping-the-first-x-lines-in-bash<\/li>
- https:\/\/unix.stackexchange.com\/questions\/349252\/cp-after-xargs-not-working<\/li><\/ul>\n","protected":false},"excerpt":{"rendered":"
Let’s say you have a JSON file and in it is an array of values. Each value has a few properties you need to pick from to ultimately create a file name. Here is an example using bash, jq, in2csv and xargs that can be strung together in one line…<\/p>\n