AmVerse Tag- and Text-review Checklist (version 2.2)

REVISION HISTORY

  1. The document should validate.
    (In A/E, use "Special/Validate document", making sure that no text is highlighted [otherwise it will validate only the selected text, not the whole document.])

  2. The structure of the document (as expressed in <FRONT>, <BODY>, <BACK>, <DIV>s of various levels and TYPEs, <LG>s, <P>s, <NOTE>s, etc.) should correspond to that of the original. <HEAD>s should appear at the head of <DIV>s.
    (In A/E, use "View/View Structural view" (hotkey: F11) to expand and contract the <DIV> structure as in an outliner. This frequently helps diagnose general structural problems.)

  3. All the <DIV>s and <LG>s should have explicit "TYPE" attributes assigned.

    (See FAQ #4 on types of LG and DIV)

  4. <PB>s should appear in the correct places, should all be present, and should accurately reflect what's printed on the page.

    5. (See the FAQ sheets for page break location and numbering.)
  5. Titlepages should be correctly tagged and numbered. Both front and back should be included.

    (See FAQ sheet #6.)

  6. In general, <NOTE>s should appear within the text at the spot to which they apply and extraneous note markers should be removed. In particular: footnotes should be embedded in the text at the point to which they refer, their "flag" characters removed from both text and note and moved to become the value of the "N" attribute of the <NOTE> tag. Long endnotes should be tagged as ordinary text and linked to the passages to which they refer using <PTR> or <REF> tags.

    See AmVerse FAQ no. 9

  7. Stuff marking the beginning and end of divisions should be correctly tagged, especially <HEAD>s (TYPE="main" and TYPE="sub" if necessary), <EPIGRAPH>s, <OPENER>s, <CLOSER>s, <TRAILER>s, and <ARGUMENT>s.

  8. If text is drama, specific drama tags should be checked.

  9. If text is poetry, indentation should be recorded using REND="ind" [for indented], REND="2ind" [for even more indented], etc. (unless the indent simply marks the beginning of a <LG>, in which case the <LG> tag itself is sufficient indication). Lines broken in two in order to fit on the page should be rejoined into one <L>.

  10. Block quotes in prose, or lines of verse quoted in prose, or separate works (poems, letters, documents) embedded in prose, should be set off with <Q> tags.

  11. Text in non-roman alphabets should be recorded with an empty GAP tag within FOREIGN tags: (<FOREIGN><GAP></FOREIGN>). Identify the language, if you know it, using the LANG attribute of the FOREIGN tag.

    See AmVerse FAQ 10 for instructions on using LANG attributes, which take an IDREF as a value and therefore require a corresponding ID in the document. We put the corresponding ID into the <HEADER>.

  12. The location of illustrations should be noted with a <FIGURE> tag, and any attached captions recorded with a <HEAD> tag within the <FIGURE> tag.

  13. Highlighted text (usually bold or italic) should be recorded with the <HI> tag (<HI REND="i"> for italic, <HI REND="b"> for bold), unless the highlighted nature of the text is being used as a cue to its structural role. E.g., if a heading is in bold, record it as a heading <HEAD>, not as a bold heading <HEAD><HI REND="b">--unless only part of it is in bold, in which case the boldness doesn't really correspond with the <HEAD>. Got that? Highlighted text usually only needs a spot check, rather than a complete page-by-page check. If in doubt, leave it in; if it's not there, put it in.

  14. Character-accuracy checks:
    1. Check that mdashes are all present and accounted for (and not recorded as hyphens). Lines ending in dash often seem to have a misplaced line tag, so check those too.

    2. Check for 0/O and 1/l confusion.

    3. Search for the "tilde" character (~) which usually means something the OCR couldn't interpret.

    4. Fix any missing or incorrect characters noticed while doing tag review. Globally check characters observed to be troublesome in the text in hand, if possible. Use a spell-checker if that helps.

    5. Proofread the whole FRONT matter.

    6. Proofread the sample pages indicated on the sheet included with the printout.

    7. If (d), (e), and (f) indicate a systematic problem of some sort, fix that problem throughout the file.

    8. If (d), (e), and (f) indicate generally poor character accuracy, proofread the entire book.

    9. If (d), (e), and (f) indicate good quality (only one character or so wrong in every 5 pages of sample), do nothing more.

  15. Claim responsibility in the document. Add something like this to the TEIHEADER, inserting a <REVISIONDESC> element at the very end of the <TEIHEADER>, using your name, the current date, and indicating in "ITEM" any change that you have made to the file. If the file required only routine corrections, as outlined in this checklist, something simple like the following will do fine. If the file had any extraordinary problems, note those and what you did to fix them (e.g. changed <NOTE>s to <HEAD>s, etc.).

    <REVISIONDESC>

    <CHANGE>
    <DATE>8/00</DATE>
    <RESPSTMT>
    <NAME>pfs</NAME>
    <RESP>MURP</RESP>
    </RESPSTMT>
    <ITEM>Proofread front matter and five-page sample. </ITEM> </CHANGE>

    <CHANGE>
    <DATE>8/00</DATE>
    <RESPSTMT>
    <NAME>pfs</NAME>
    <RESP>MURP</RESP>
    </RESPSTMT>
    <ITEM>Corrected and revised markup.</ITEM>
    </CHANGE>

    </REVISIONDESC>

  16. Manage your files (see FAQ sheet #8. Make sure especially that:

Index