bisac.csv

bisac.go

// Package bisac provides a mapping of BISAC codes to their descriptions.
package bisac

import (
	"bytes"
	_ "embed"
	"encoding/csv"
	"fmt"
	"io"
	"strings"
)

//go:embed bisac.csv
var data []byte

type Code string

func InitCodes() {
	codes()
}

func (c Code) String() string {
	return string(c)
}

func (c Code) Description() string {
	return codes()[string(c)]
}

func (c Code) Valid() bool {
	_, ok := codes()[string(c)]
	return ok
}

var cache map[string]string

func codes() map[string]string {
	if cache == nil {
		cache = make(map[string]string)
		r := csv.NewReader(bytes.NewReader(data))
		for i := 0; true; i++ {
			x, err := r.Read()
			if err == io.EOF {
				break
			}
			if err != nil {
				panic(fmt.Errorf("bisac: parse embedded csv: %w", err))
			}
			if i == 0 {
				if len(x) < 1 || x[0] != "Code" {
					panic(fmt.Errorf("bisac: parse embedded csv: incorrect header %+q: column 1 is not code", x))
				}
				if len(x) < 2 || x[1] != "Description" {
					panic(fmt.Errorf("bisac: parse embedded csv: incorrect header %+q: column 2 is not description", x))
				}
				if len(x) < 3 || x[2] != "Comment" {
					panic(fmt.Errorf("bisac: parse embedded csv: incorrect header %+q: column 3 is not comment", x))
				}
				if len(x) >= 4 {
					panic(fmt.Errorf("bisac: parse embedded csv: incorrect header %+q: too many fields", x))
				}
			} else if i == 1 && x[0] == "" {
				// ignore
			} else {
				if len(x[0]) != 9 {
					panic(fmt.Errorf("bisac: parse embedded csv: row %+q: code not 9 characters", x))
				}
				if !strings.Contains(x[1], " / ") {
					panic(fmt.Errorf("bisac: parse embedded csv: row %+q: description does not contain slash", x))
				}
				cache[x[0]] = x[1]
			}
		}
	}
	return cache
}

scrape.js

// https://bisg.org/page/bisacedition
await (async function() {
    async function fetchHTML(url) {
        const resp = await fetch(url)
        if (resp.status != 200) {
            throw new Error(`response status ${resp.status}`)
        }
        return new DOMParser().parseFromString(await resp.text(), "text/html")
    }

    function toTitleCase(str) {
        return str.replace(/\w\S*/g, w => w.charAt(0).toUpperCase() + w.substr(1).toLowerCase())
    }

    function isBISAC(str) {
        return /^[A-Z]{3}[0-9]{6}$/.test(str)
    }

    function formatCSV(arr) {
        return arr.map(x => x.map(y => JSON.stringify(y)).join(",")).join("\n") + "\n"
    }

    function elideSpaces(str) {
        return str.trim().split(/\s+/).join(" ")
    }

    function saveFile(fn, txt) {
        const a = document.createElement("a")
        a.download = fn
        a.href = "data:application/octet-stream;base64," + btoa(txt)
        a.click()
    }

    const codes = []
    for (const a of document.getElementById("CustomPageBody").querySelectorAll("ul>li a[href*='/page/']")) {
        if (a.innerText.trim().length == 0) {
            continue
        }

        const url = a.href.toString().replace(/^http:/, "https:")
        console.log(`[scrape] ${elideSpaces(a.innerText)} - ${url}`)

        const doc = await fetchHTML(url)
        for (const p of doc.getElementById("CustomPageBody").querySelectorAll("p")) {
            const spl = elideSpaces(p.innerText).split(" ")
            if (spl.length >= 2 && isBISAC(spl[0])) {
                const code = spl.shift()
                const labelEnd = spl.findIndex(x => /\(|\*/.test(x))
                const comment = labelEnd != -1 ? spl.splice(labelEnd).join(" ") : null
                const labelSpl = spl.join(" ").split("/")
                const label = toTitleCase(labelSpl.shift()) + "/" + labelSpl.join("/")
                codes.push([code, label, comment ? comment : ""])
            }
        }
    }

    saveFile("bisac.csv", formatCSV([
        ["Code", "Description", "Comment"],
        ["", elideSpaces(document.getElementById("SpTitleBar").innerText), `Generated on ${new Date().toISOString()} from ${window.location}`],
        ...codes,
    ]))

    return codes
})()

How to run this code? @pgaskin