0.6 Beta take 2 (0.5.905)

[mrmikel]

I guess I mean macro. I know I can change assignments, but I would like to add to them. I would like to be able to hit F7 and have it do space then delete (to join paragraphs together that are not easily done with Regex.) I personally would only need a small number of actions, perhaps no more than 5, since you could always do one after the other.

Thanks for the other key strokes. Will give it a whirl.

[Serpentine]

Or just use any macro application?

[kiwidude]

@Serpentine - one of the changes that was made with this beta build was to use a different third party zip library. It appears that this library Sigil now uses is not detecting the container.xml file in that zip file, so it doesn't extract it. As Sigil is hard-coded to "expect" a META-INF/container.xml file to be present in an EPUB, it tries to grab the file from the expected location it had unzipped to, and doesn't find it, hence the error message that gets displayed.

Did you happen to package the EPUB yourself (if so how?). I'm wondering what tool earlier in the chain has got this file into this state. At the same time we can look into both perhaps why the zip library cannot detect the file (but other tools like winzip can) and perhaps a more elegant/meaningful message if the file genuinely does not exist.

[meme]

Originally Posted by Toxaris

Saw the index function, great! Very easy in usage. Not that I will use it that much, but still...

Yeah, not that necessary but its available when it is. The autofill option is always worth a try

[Serpentine]

Originally Posted by kiwidude

Did you happen to package the EPUB yourself (if so how?). I'm wondering what tool earlier in the chain has got this file into this state. At the same time we can look into both perhaps why the zip library cannot detect the file (but other tools like winzip can) and perhaps a more elegant/meaningful message if the file genuinely does not exist.

Sadly it wasn't packaged by me. I'll have a look and see if I can figure anything more out about the archive. I've used minizip in the past but only the classic version (there are two sort-of stable versions in use - both have some questionable code). Never had issues with it tho (but swapped out for libarchive since the minizip api changes were annoying).

[kiwidude]

@Serpentine - you did open it in calibre yesterday though right I see from the calibre_bookmarks.txt in there . I have a suspicion calibre is causing the "problem", but I need to turn back on bookmarks in my viewer etc to confirm if that is the case. Regardless we will put in place something better for the next build than the nasty error you get now.

[Serpentine]

Originally Posted by kiwidude

@Serpentine - you did open it in calibre yesterday though right I see from the calibre_bookmarks.txt in there

Yeah, I wanted to make sure it was a sigil issue and not related to the file itself. However calibre does not cause the problem, the unchanged file still has the same issue. Adding to a zip doesn't rewrite the archive

[Man Eating Duck]

Originally Posted by meme

The autofill option is always worth a try

Yeah, that's... helpful

We tend to omit indices in epubs, even when the DT editions has one. They're a bit superfluous when all readers and software have search functions. If you want to make the index functionality more useful, however, I have a couple of suggestions:

• Have separate columns for display and hit words; ie for the entry "vehicle" you could have it generate hits for "bus", "car", "boat" and so on
• Case insensitivity in the generation code. Capitalised words could occur at the start of a sentence.
• Have it pick up a marker such as a span class instead of all search hits. If England is mentioned many times in a few paragraphs, you really only need a link for the first one. This might be automated from the UI with a "Mark selected for index" button.
• Possibly make it multilevel. I don't really know how this should be handled in the UI, though.

As I said, at the end of the day I don't believe indexes will be necessary as a lookup tool, but they might be helpful as a "Summary of concepts" touched upon in the book.

On a side note I've been using the beta quite a bit now, and even prepared a commercial book for publication with it (might it be the first one made with the beta?). I love it, you guys have done an outstanding job.

[meme]

Originally Posted by Man Eating Duck

Yeah, that's... helpful

We tend to omit indices in epubs, even when the DT editions has one. They're a bit superfluous when all readers and software have search functions. If you want to make the index functionality more useful, however, I have a couple of suggestions:

• Have separate columns for display and hit words; ie for the entry "vehicle" you could have it generate hits for "bus", "car", "boat" and so on
• Case insensitivity in the generation code. Capitalised words could occur at the start of a sentence.
• Have it pick up a marker such as a span class instead of all search hits. If England is mentioned many times in a few paragraphs, you really only need a link for the first one. This might be automated from the UI with a "Mark selected for index" button.
• Possibly make it multilevel. I don't really know how this should be handled in the UI, though.

As I said, at the end of the day I don't believe indexes will be necessary as a lookup tool, but they might be helpful as a "Summary of concepts" touched upon in the book.

On a side note I've been using the beta quite a bit now, and even prepared a commercial book for publication with it (might it be the first one made with the beta?). I love it, you guys have done an outstanding job.

The user guide's a work in progress - and would be a lot easier if we didn't keep changing everything every day

Yes, I know an index isn't needed by most, but a certain person around here once asked about it and it seemed interesting to see what could be done...

• There are 2 columns - one for the text to match, and one for the text to display - with the display defaulting to the text if you don't enter anything
• I think I made it case sensitive on purpose in case of proper names, etc. Just use Brown, brown to pick up both (tooltip should mention this)
• Use Mark for Index on your text and you get a span marker
• It is multilevel - just use First/Second in the text to use in the index column

Yeah, it is great, isn't it

[Man Eating Duck]

Originally Posted by meme

The user guide's a work in progress - and would be a lot easier if we didn't keep changing everything every day

Yes, I know an index isn't needed by most, but a certain person around here once asked about it and it seemed interesting to see what could be done...

• There are 2 columns - one for the text to match, and one for the text to display - with the display defaulting to the text if you don't enter anything
• I think I made it case sensitive on purpose in case of proper names, etc. Just use Brown, brown to pick up both (tooltip should mention this)
• Use Mark for Index on your text and you get a span marker
• It is multilevel - just use First/Second in the text to use in the index column

Yeah, it is great, isn't it

Obviously I didn't play enough with it, but it's amazing that I described almost the complete feature set... as suggestions. Shall I consider it a sign that great minds think alike?

[JMikeD]

Originally Posted by kiwidude

I will say that out of the approaching 4,000 downloads of this Sigil beta, other than some MacOS users as per above you are the one and only user who has reported a crash problem with this Sigil build.

Just tried 0.5.905 on my MacBook with OS 10.7.5. Sigil crashed immediately.

Process: launchd [8650]
Path: /Applications/Sigil.app/Contents/MacOS/Sigil
Identifier: com.sigil-ebook.Sigil.app
Version: ??? (???)
Code Type: X86-64 (Native)
Parent Process: launchd [333]

Date/Time: 2012-09-27 17:19:55.243 -0500
OS Version: Mac OS X 10.7.5 (11G56)
Report Version: 9

Interval Since Last Report: 373922 sec
Crashes Since Last Report: 73
Per-App Crashes Since Last Report: 3
Anonymous UUID: 835CA56E-3F8B-4832-9A89-6558FFC1E844

Crashed Thread: Unknown

Exception Type: EXC_BAD_ACCESS (SIGSEGV)
Exception Codes: KERN_INVALID_ADDRESS at 0x00007fff5fc01028

Backtrace not available

Unknown thread crashed with X86 Thread State (64-bit):
rax: 0x0000000000000055 rbx: 0x0000000000000000 rcx: 0x0000000000000000 rdx: 0x0000000000000000
rdi: 0x0000000000000000 rsi: 0x0000000000000000 rbp: 0x0000000000000000 rsp: 0x0000000000000000
r8: 0x0000000000000000 r9: 0x0000000000000000 r10: 0x0000000000000000 r11: 0x0000000000000000
r12: 0x0000000000000000 r13: 0x0000000000000000 r14: 0x0000000000000000 r15: 0x0000000000000000
rip: 0x00007fff5fc01028 rfl: 0x0000000000010203 cr2: 0x00007fff5fc01028
Logical CPU: 1

Binary images description not available

I've been using 0.5.3 previously, no problem.

[kiwidude]

@JMikeD - see the very first post and the previous page - there is a known packaging issue with the current beta for some MacOS users. It is intended to be fixed for the next release.

[kiwidude]

Originally Posted by Serpentine

Yeah, I wanted to make sure it was a sigil issue and not related to the file itself. However calibre does not cause the problem, the unchanged file still has the same issue. Adding to a zip doesn't rewrite the archive

Fixed for the next release - the issue was the epub did not have mimetype as the first file in it, and a subtle bug in Sigil meant the first file was accidentally being skipped. So if that first file it bypassed happened to be container.xml...

[Man Eating Duck]

An improvement request: I de-obfuscate embedded fonts as I only use freely licenced fonts, which Indesign unhelpfully obfuscates. Is there any possibility of adding a way of setting obfuscation options for all fonts in the epub at once? When selecting multiple fonts the options are grayed out even if all fonts have the same obfuscation settings. In my case it might also be a preference option. And I consulted the user guide this time, honestly

@kiwidude: Font de-obfuscation might also be a useful option for Modify epub, if Calibre has an API for it? I couldn't find one, or I would have given it a go myself.

[kiwidude]

@Man Eating Duck - It sounds perfectly reasonable on the surface. I suspect the reason why the menu is greyed out is purely the complexity of dealing with the submenu with a multiple selection made. For instance if you selected two fonts, one was obfuscated, and the other wasn't. Or each had different obfuscation schemes. I've put it on the list to be looked at sometime...