2b:null 2c:["$","title",null,{"children":"Collecting Sol LeWitt's Prompts into a Data Set | Tanner Hobson | Observable"}] 31:T4bb,# Collecting Sol LeWitt's Prompts into a Data Set I saw a blog post the other day about using [ChatGPT with GPT-3.5](https://www.amygoodchild.com/blog/chatgpt-sol-lewitt-wall-drawings) (and later [ChatGPT with GPT-4](https://www.amygoodchild.com/blog/ai-generations-chatgpt-4-sol-lewitt-wall-drawings)) to interpret [Sol LeWitt's "Wall Drawing" series](https://en.wikipedia.org/wiki/Sol_LeWitt#Wall_drawings) and "see" what the AI thinks the prompts mean and how it tries to follow the instructions. I was inspired by this and wanted to find a way to use ChatGPT to create new Wall Drawings that I could follow myself. This notebook does not yet solve that problem. What this notebook does solve, is the problem of collecting the Wall Drawings prompts in some machine readable format. I found [a website on Sol LeWitt](https://massmoca.org/sol-lewitt/) that includes a good selection of prompts. Using that website as a source, I manually copy and pasted a lot of the pages and prompts into a set of text files and uploaded it here for further processing (under the name `Sol LeWitt.zip`). This notebook includes code to process that zip file and turn it into another zip file that is even more structured.32:T5be,{ const z = zip(); for (const extract of _EXTRACTS_FINAL) { const directory = DIRECTORY_FOR({ extract }); EXPORT_AS_TEXT({ extract, property: 'text', filename: `${directory}/OriginalPrompt.txt` }); EXPORT_AS_TEXT({ extract, property: 'prompt', filename: `${directory}/Prompt.txt` }); EXPORT_AS_TEXT({ extract, property: 'date', filename: `${directory}/Date.txt` }); EXPORT_AS_TEXT({ extract, property: 'medium', filename: `${directory}/Medium.txt` }); EXPORT_AS_TEXT({ extract, property: 'discussion', filename: `${directory}/Discussion.txt` }); EXPORT_AS_TEXT({ extract, property: 'backstory', filename: `${directory}/Backstory.txt` }); z.file(`${directory}/Metadata.json`, JSON.stringify({ name: extract.name, text: extract.text, prompt: extract.prompt, date: extract.date, medium: extract.medium, discussion: extract.discussion, backstory: extract.backstory, entities: Array.from(extract.entities).reduce((acc, [key, val]) => { acc[key] = val; return acc; }, {}), }, true, 2)); } const blob = await z.generateAsync({ type: 'blob' }); return DOM.download(blob, `Sol LeWitt Data Set.zip`, `Download Sol LeWitt Data Set.zip`); function DIRECTORY_FOR({ extract }) { const { filename } = extract; return filename.split('.', 2)[0]; } function EXPORT_AS_TEXT({ extract, property, filename }) { z.file(filename, extract[property]); } }33:Ta41,async function EXTRACT(filename) { const name = filename.split('/')[1].split('.')[0]; const number = name.split('Wall Drawing ', 2)[1]; const file = solLewitt1.file(filename); let text = await file.text(); for (const x of ['', '', '', '', '', '\r']) { text = text.replaceAll(x, ''); } const originalText = text; let entities = new Map(); const entityRegex = []; entityRegex.push( new RegExp(String.raw`_(Wall Drawing ${number}[^_]*)_`/* String.raw */, 'g') ); entityRegex.push( /(Sol LeWitt)/g ); entityRegex.push( /(LeWitt)/g ); entityRegex.push( /\b([A-Z][a-zA-Z]+(?:\s+[A-Z][a-zA-Z]+)+)\b/g ); entityRegex.push( /_([^_]+)_/g ); for (const re of entityRegex) { for (const match of text.matchAll(re)) { const entity = match[1]; if (entity === 'Wall Drawings') continue; if (!entities.has(entity)) { entities.set(entity, `$${entities.size+1}`); } } } for (const [entity, replacement] of entities.entries()) { text = text.replaceAll(`_${entity}_`, replacement); text = text.replaceAll(`${entity}`, replacement); } entities = new Map(Array.from(entities.entries()).map(([k, v]) => ([v, k]))); let backstory; [text, backstory] = text.split('**Backstory**', 2); let lines = text.split('\n\n'); // lines = lines.filter((d) => !d.includes('- Sol LeWitt')); // lines = lines.filter((d) => !d.includes('Yale University')); // lines = lines.filter((d) => !d.includes('First Installation')); // lines = lines.filter((d) => !d.includes('First Drawn By')); // lines = lines.filter((d) => !d.includes('Collection ')); // lines = lines.filter((d) => !d.includes(' Collection')); // lines = lines.filter((d) => !d.includes('MoCA')); const prompt = lines[0].trim(); const date = lines[1].trim(); const medium = lines[2].trim(); const discussion = lines.slice(3).join('\n\n').trim(); backstory = (backstory || '').trim(); return Object.assign({ entities, name, filename, prompt, date, medium, discussion, backstory, lines, text: originalText }, { view() { return htl.html`

${name}

${d}: ${d==='entities' ? Array.from(entities.entries()).map(([k, v]) => `${k}: ${v}`).join('\n') : this[d]} `/* htl.html.fragment */)} <dt>Inspect <dd>${inspect(this)} `/* htl.html */; }, }); }34:[{"id":"9dd065947ec0bc9a597afafa219b455ddb91806fd8eff2dd25ea69c7ee2e3ca1a101f627684a5bf69979f06ddb49aa6198c2bd1202fe5b62d7d14af14d0aa662","href":"https://static.observableusercontent.com/files/9dd065947ec0bc9a597afafa219b455ddb91806fd8eff2dd25ea69c7ee2e3ca1a101f627684a5bf69979f06ddb49aa6198c2bd1202fe5b62d7d14af14d0aa662","name":"Sol LeWitt@2.zip","type":"application/zip","lastModified":1679775438819.097,"size":42705},{"id":"679375abf8d5cb90e691c076ce64011166ce10f03563ba3c7c2bc1552f618d0c0f282855d68a0aa7503561e20c3b19d24f29b13d0eaa2b1830126d745b984404","href":"https://static.observableusercontent.com/files/679375abf8d5cb90e691c076ce64011166ce10f03563ba3c7c2bc1552f618d0c0f282855d68a0aa7503561e20c3b19d24f29b13d0eaa2b1830126d745b984404","name":"Sol LeWitt@1.zip","type":"application/zip","lastModified":1679763137504.827,"size":51496},{"id":"03debd317f4354880912c4520b14cc35785c0dfdb1cc0d7d66c3f97dc0b7add16b38b6cc8d6a448c058c46494ddc01670784c8a6faa24d84423490e9560cb3b0","href":"https://static.observableusercontent.com/files/03debd317f4354880912c4520b14cc35785c0dfdb1cc0d7d66c3f97dc0b7add16b38b6cc8d6a448c058c46494ddc01670784c8a6faa24d84423490e9560cb3b0","name":"Sol LeWitt Data Set.zip","type":"application/zip","lastModified":1680551978716.324,"size":354882}] 2f:{"id":"356e0bd34b0c918d","slug":null,"title":"Collecting Sol LeWitt's Prompts into a Data Set","thumbnail":"77e22d5deafc9703b3971adcbc3272dc6d492f87a809371c80f693bd4fcb095b","access_level":"public","create_time":"$D2023-03-25T04:48:45.851Z","update_time":"$D2023-04-03T19:58:41.731Z","view_time":null,"first_public_time":"$D2023-03-25T04:48:45.851Z","fork":false,"stars":0,"starred":false,"owner":{"type":"team","id":"7343f790779e6735","login":"player1537","name":"Tanner Hobson","avatar_url":"https://avatars-next.observableusercontent.com/16170a1a32e364e1ff6054c6045de6814a032d01331aba01288f82afceeaaec2","tier":"starter_2024"},"version":143,"publish_version":null,"role":null,"authored":false,"authors":[{"type":"individual","id":"e0a627f6b6f3beb0","login":"player1537","name":"Tanner Hobson","avatar_url":"https://avatars-next.observableusercontent.com/16170a1a32e364e1ff6054c6045de6814a032d01331aba01288f82afceeaaec2","tier":null}],"members":[{"type":"individual","id":"e0a627f6b6f3beb0","login":"player1537","name":"Tanner Hobson","avatar_url":"https://avatars-next.observableusercontent.com/16170a1a32e364e1ff6054c6045de6814a032d01331aba01288f82afceeaaec2","tier":null,"role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"lapsed":null,"subscribed":null,"forks":1,"collection":{"id":"d21bf020f9cae8af","slug":"work-in-progress-promising","title":"Work-in-Progress (promising)","owner":{"id":"7343f790779e6735","login":"player1537","name":"Tanner Hobson"}},"type":"notebook","body":{"title":"Collecting Sol LeWitt's Prompts into a Data Set","cells":[{"id":0,"value":"$31","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":122,"value":"$32","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":140,"value":"{\n const extracts = _EXTRACTS_FINAL;\n return Inputs.textarea({ rows: 24, cols: 24, value: extracts.map(({ prompt }) => prompt).join('\\n\\n') })\n}","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":102,"value":"mutable _EXTRACTS_FINAL = null","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":106,"value":"{\n const iter = EXTRACTS();\n for (;;) {\n const { value, done } = await iter.next();\n if (!done) {\n yield value;\n } else {\n mutable _EXTRACTS_FINAL = value;\n break;\n }\n }\n}","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":7,"value":"async function *EXTRACTS() {\n const details = htl.html`<details>`;\n \n const extracts = [];\n yield VIEW();\n\n let filenames = solLewitt1.filenames;\n filenames = [...filenames];\n filenames.sort((a, b) => {\n a = +a.match(/\\d+/)[0];\n b = +b.match(/\\d+/)[0];\n return a - b;\n });\n \n for (let i=0, n=filenames.length; i<n; ++i) {\n const filename = filenames[i];\n const extract = await EXTRACT(filename);\n extracts.push(extract);\n yield VIEW();\n }\n return extracts;\n\n function VIEW() {\n details.replaceChildren(htl.html.fragment`\n <summary>${extracts.length} drawings</summary>\n ${extracts.map((extract) => extract.view())}\n `/* htl.html.fragment */);\n return details;\n }\n}","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":28,"value":"$33","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":5,"value":"solLewitt1 = FileAttachment(\"Sol LeWitt@2.zip\").zip()","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":60,"value":"import {inspect} from \"@observablehq/inspector\"","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":118,"value":"import {zip} from \"@fil/jszip\"","pinned":true,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"}],"files":"$W34","stdlib":"1","titling":"auto"}} 30:"$2f:body" </article></body></html>