; page.sol -- read an HTML file and report on it.
;
; Run with: ./bin/solas programs/page.sol && ./bin/solvm programs/page.sob
; Over a file of your own: ./bin/solvm programs/page.sob path/to/page.html
; Listing one kind of element: ./bin/solvm programs/page.sob page.html img
;
; The parser is [lib/html.sol](../lib/html.sol), on the search path, so this
; says `@include "html.sol".` and not where it lives.
;
; The fourth program here, and the one that pushed on ground the other three
; did not:
;
; 1. **It cannot fail on bad input, because bad input is the normal case.**
; log.sol skips a bad line, json.sol refuses a bad document -- both right
; for their format. HTML is generated, served, and wrong, so a reader that
; stops is no use. The parser recovers and keeps a list, and this program
; prints the list as part of the report rather than as an error.
;
; 2. **The tree is built against a stack, not by recursion**, and that is what
; takes it out of reach of the 62-frame limit (ROADMAP 3.5) that stops
; json.sol at 124 levels of nesting. Measured: this reads a document nested
; 50,000 deep. The catch was that walking the tree back down recursed, and
; capped at 28 again -- so `text`, `find` and `findAll` are written with a
; stack too. The limit is not a property of the data; it is a property of
; how you traverse it.
;
; 3. **An array could not be popped, or asked what it holds**, which a stack
; notices immediately. Both were written around here first and both are
; built now -- `removeLast` and `indexOf`, COMPLETED 6.23. The workaround
; being in shipped library code is what made the case for them.
@include "html.sol".
; ---------------------------------------------------------------------------
; Where the page comes from
sample := "
Solveig — a small language
Solveig
A small object-oriented language, its bytecode virtual machine, and a REPL.
Everything is an object & all work happens by sending messages.
Getting started
Examples
There are 27 of them, and three are whole programs.
![\"the](\"diagram.png\")
![](\"screenshot.png\")
Status
Version 0.4.0.
".
path := system:arguments:size:greaterThan(#0):ifElse(
{ system:arguments:at(#1) },
{ | fallback |
fallback := "build/example.html".
system:writeFile(fallback, sample).
fallback }).
system:fileExists(path):ifFalse({
"no such file: {}":fill([path]):display.
system:exit(#1) }).
; No `onError` around this one, and that is the point: it does not raise.
page := html:read(system:readFile(path)).
; ---------------------------------------------------------------------------
; What is in it
title := page:find("title").
"":display.
"{} -- {} bytes":fill([path, system:fileSize(path)]):display.
title:isNil:ifElse(
{ "no
":display },
{ "title: {}":fill([title:text]):display }).
; A tag named on the command line is dumped and nothing else is, which is what
; makes this usable as a tool rather than only as a demonstration.
system:arguments:size:greaterThan(#1):ifTrue({ | wanted, found |
wanted := system:arguments:at(#2):asLowercase.
found := page:findAll(wanted).
"":display.
"{} <{}>":fill([found:size, wanted]):display.
found:do({ e |
" {}{}":fill([
e:attributes:size:equals(#0):ifElse({ "" }, {
e:attributes:keys:sorted:collect({ k |
"{}=\"{}\"":fill([k, e:attributes:at(k)]) }):join(" "):concat(" ") }),
e:text]):display }).
system:exit(#0) }).
; ---------------------------------------------------------------------------
; The outline
;
; Every heading, in document order, which one walk gives and six calls to
; `findAll` would not -- those would answer all the h1s, then all the h2s.
; `selectNodes` takes the test as a block, so "is this a heading" is written
; once and the order is the document's.
isHeading := { node | "h1 h2 h3 h4 h5 h6":indexOf(node:name):notNil }.
headings := page:selectNodes(isHeading).
"":display.
headings:size:equals(#0):ifElse(
{ "no headings":display },
{ "outline":display.
headings:do({ h | | level, indent |
level := h:name:copyFrom(#2, #2):asInteger.
indent := "".
[#1,level:sub(#1)]:loop({ n | indent := indent:concat(" ") }).
" {}{}":fill([indent, h:text]):display }) }).
; ---------------------------------------------------------------------------
; Links and images
links := page:findAll("a").
"":display.
"{} links":fill([links:size]):display.
links:do({ a | | href |
href := a:attribute("href").
" {} {}":fill([
href:isNil:ifElse({ "(no href)" }, { href:asString("<22") }),
a:text]):display }).
; The one check worth making on a page, and the reason `attribute` answers nil
; rather than raising: a missing attribute is a question, not a failure.
images := page:findAll("img").
missing := images:select({ img | img:attribute("alt"):isNil }).
"":display.
"{} images, {} without alt text":fill([images:size, missing:size]):display.
missing:do({ img |
" {}":fill([img:attribute("src"):isNil:ifElse(
{ "(no src)" }, { img:attribute("src") })]):display }).
; ---------------------------------------------------------------------------
; What was wrong with it
;
; Printed as part of the report rather than as an error, because for this format
; they are a finding about the page and not a failure of the reader. The tree
; above was built from the same document that produced them.
"":display.
html:complaints:size:equals(#0):ifElse(
{ "no complaints":display },
{ "{} thing{} wrong with the markup":fill([html:complaints:size,
html:complaints:size:equals(#1):ifElse({ "" }, { "s" })]):display.
html:complaints:do({ c | " ":concat(c):display }) }).