0.5.2 Released

[Hitch]

@st_albert: I'm not actually known for a surfeit of patience. Being one of the naggers who asked for PCRE, I'm trying, I really am, to be patient. But the Dodo thing is rather vexing.

@mmat1: Well, "replace all" would sucketh in the circumstances I'm looking at...I'll either have to regex it, to avoid instances in which an alpha comes after, (which now scares the big-girl panties right off of me), or do it manually, which is starting to make me cranky. (Those who know me would tell you that they can't notice a difference in the level of cranky, so it mayn't matter).

And, thanks, Toxaris. ;-)

And, while we're at it, if we're adding WP-like stuff, why not do something REALLY bloody useful for an ePUB, and add a Concordancer? I mean, seriously....as a pro, we get a bajillion books in here that have indices. I can't convince the authors to abandon the indices, naturally, no matter how many times I explain about "search." We are therefore left with either a solution in which the previously-linked (top-of-page) links no longer "find" the referenced word (naturally), or we have to do a concordance-like index by hand. Given the nature of ebooks, a concordance-like index would actually make sense (giving the publisher the choice of which to leave in and which to omit), if you could auto-build the links while you were at it, across Sigil.

Just a thought. Whilst I'm being cranky...

[DiapDealer]

The F&R works fine for me in 0.5.1. As does Spell check.

There's that niggling little crash in 0.5.1 when trying to add existing files (images, fonts, etc..) from the context menu (right clicking on files), but that's easy enough to work around by using the File->New->Add Existing File(s) menu.

[Toxaris]

Originally Posted by mmat1

There are 3 possible workarounds:
1. bravely hit "replace all", this schould work always
2. Install 5.1. There are some other issues on 5.1, which maybe won't hurt you. S/R and spellcheck should work.
3. I've seen a patch at Sigil's web site (don't ask me how to handle this, but it's there)

Of course, there are also problems with these work arounds. Often I want to search a certain passage and not a replace at all. So, option 1 does not cut it.
For option 2 I have other issues, namely frequent crashes at S&R activities. Also not very handy.
Option 3 I will not do. I don't need all kind of build tools and don't want to build myself. That I will leave to people who know what to do.

The solution I work with currently, is to open my source HTML in Notepad++ and execute the searches there. The line number gives an indication where to look in Sigil. Not ideal, but for now it works as long as I don't split in Sigil

[mmat1]

Originally Posted by Toxaris

Of course, there are also problems with these work arounds.

I know that my suggestions would not lead to a perfect solution, as well as the fourth one:

- turning search mode to "current file"

I'm just looking for an idea how to get things done under certain circumstances.

[davidfor]

Originally Posted by Hitch

@mmat1: Well, "replace all" would sucketh in the circumstances I'm looking at...I'll either have to regex it, to avoid instances in which an alpha comes after, (which now scares the big-girl panties right off of me), or do it manually, which is starting to make me cranky. (Those who know me would tell you that they can't notice a difference in the level of cranky, so it mayn't matter).

One thing people missed mentioning is that the find-and-replace is only failing when you search all files. If you only search the current file, it works correctly. What I found I had to do when fixing an epub was to do all the changes in one file, open the next file and do a search all files from there to find the next occurrence and then change back to just the current file to fix this. It is a pain but it works. And I will be very happy to see 0.5.3 with this fixed.

[meme]

Originally Posted by Hitch

(P.S. -- if we are indeed persisting in adding a spell-checker to Sigil, as I see we are, can we at least add finding the next instance of a purportedly misspelled word to it, so I don't have to sit here and scroll? I feel like I've flashed back in time 20 years. If we're going to have a wp-like feature, can we make it a skosh more usable? Thx. Sorry if I sound cranky.)

Originally Posted by Hitch

@st_albert: I'm not actually known for a surfeit of patience. Being one of the naggers who asked for PCRE, I'm trying, I really am, to be patient. But the Dodo thing is rather vexing.

Originally Posted by Hitch

And, while we're at it, if we're adding WP-like stuff, why not do something REALLY bloody useful for an ePUB, and add a Concordancer? I mean, seriously....as a pro, we get a bajillion books in here that have indices. I can't convince the authors to abandon the indices, naturally, no matter how many times I explain about "search." We are therefore left with either a solution in which the previously-linked (top-of-page) links no longer "find" the referenced word (naturally), or we have to do a concordance-like index by hand. Given the nature of ebooks, a concordance-like index would actually make sense (giving the publisher the choice of which to leave in and which to omit), if you could auto-build the links while you were at it, across Sigil.

0.5.3 will fix the Find issue and shouldn't be too long now.

As for spell check, I think that's something a lot of people want - just a question of fitting it in with everything else There's no actual issue opened for it so I'll open one.

As for a a Concordance, I do see issue 167 for that, so its been around a long time. How would you see this working - not that its at the top of the list But actual details are worth documenting for when someone gets around to it. What do you mean by top-of-page links? Without having looked into indexes, I assume you're asking for Sigil to scan for every word in your document (or word pairs/word triplets...), allow you to select/unselect them in a list, and then generate a page listing them with links back to each occurrence - meaning every occurrence must have an id (well, everyone that's linked to). And since there aren't page numbers, what do you put in the index under "word" - 1.2.3.4, or chapter 5, chapter 5, chapter 7...

[Rand Brittain]

Thanks for all your hard work fixing bugs!

I was all "ha ha I can just keep using 0.5.1" and then my book was like "ha ha you gotta add some images" and I was like "noooooooooooo". So that's how it stands!

[Hitch]

Originally Posted by meme

0.5.3 will fix the Find issue and shouldn't be too long now.

As for spell check, I think that's something a lot of people want - just a question of fitting it in with everything else There's no actual issue opened for it so I'll open one.

Thanks, and thanks. I was one of the people (and still am, honestly) who was not thrilled with the idea of adding spell-check to Sigil, as it feels very "word-processor-ish" to me, which is not a direction I--personally--really want to see Sigil continue toward. However, as it's clearly here, and here to stay, then let's make it a bit more useful, as being 1) able to go to the next incidence, and b) some sort of temporary buffer for words that we don't want to add to our dictionaries permanently, but don't want to have to scroll through/leap to for every instance, like (heavens help us all) dialect, or NAMES, or god-knows-what. I did a book today with a spelling of "Victor" that was "Viktor," which I didn't want to add to my dictionary for obvious reasons, and I don't know how many times it appeared in this masterpiece, but, lemme tell ya: it was a blooody LOT.

As for a a Concordance, I do see issue 167 for that, so its been around a long time. How would you see this working - not that its at the top of the list But actual details are worth documenting for when someone gets around to it. What do you mean by top-of-page links? Without having looked into indexes, I assume you're asking for Sigil to scan for every word in your document (or word pairs/word triplets...), allow you to select/unselect them in a list, and then generate a page listing them with links back to each occurrence - meaning every occurrence must have an id (well, everyone that's linked to). And since there aren't page numbers, what do you put in the index under "word" - 1.2.3.4, or chapter 5, chapter 5, chapter 7...

Well, what I mean by top of page links is actually misspeaking, BUT, when someone builds an index in Word (or Wordperfect, or INDD), the cross-referenced links actually TARGET the page number--the header at the top of the page (actually...the first position before the first letter of the first word of the first line at the top of the page, to be technical). They're not links to the actual word, not when the html is exported. Thus, the links (and the "original" page numbers) are output to the html that we poor converters get to convert FROM. So you end up with someone clicking "page 53" for an indexed word, and sure enough, he's taken to what used to be the first line of what used to be page 53--which now has as much relevance to the word/phrase/concept being searched as my grannie's undies. So, They Is Worthless. However, other than doing hand-linking to the actual word/phrase/concept itself, there's nothing better to be had.

So, yes, I'd think that a Concordancer that works very much as you've described, with a numbering scheme of 1, 2, 3 (or i, ii, iii--that part doesn't matter to me as much) would be pretty damned brilliant. I'd LOVE to be able to offer that to my clients in lieu of really aggravating indices that aren't worth a hoot, are frankly outmoded and are invariably a source of frustration. A concordanced index WOULD, for non-fiction and reference books, be fantastic (at least, for the ones I use, I'd love it.) I also use the search function(s) extensively, myself...but for those authors who can't let go of an index (and I can't quite blame them--I use Liz Castro's all the time), I'd like to be able to affordably build an index that's worth a damn. And what you've described sounds really quite the thing.

(sorry to ramble--long few days and tired. Very trying books/clients lately.)

Hitch

[JonathanH]

What's the best recent build to use for multi-file search-and-replace while we wait for 0.5.3? I seem to recall 0.5.1 would sometimes just crash doing search-and-replace. Should I go to a pre-5.0 version?

[theducks]

Originally Posted by JonathanH

What's the best recent build to use for multi-file search-and-replace while we wait for 0.5.3? I seem to recall 0.5.1 would sometimes just crash doing search-and-replace. Should I go to a pre-5.0 version?

I am using 5.1 on XP
I have done a lot of S&R (CV and REGEX) without a crash. I do frequent saves after each pass.

[meme]

Toxaris - you mentioned you have frequent crashes in S&R. Do you have any details on this. As far as I'm aware this should only happen in Replace step by step and only if you have a large file (and will remain until the Book View editor is replaced).

Hitch - thanks for the info on the Concordance. Its good to have wishlists

[st_albert]

Originally Posted by JonathanH

What's the best recent build to use for multi-file search-and-replace while we wait for 0.5.3? I seem to recall 0.5.1 would sometimes just crash doing search-and-replace. Should I go to a pre-5.0 version?

If you don't need spell check, and you don't need full PCRE regex, version 0.4.2 is pretty stable in my experience.

[JonathanH]

Originally Posted by st_albert

If you don't need spell check, and you don't need full PCRE regex, version 0.4.2 is pretty stable in my experience.

Reverting to 0.5.1 on one machine just to get full search has worked nicely - thanks all!