[Solved] Can dictionary get too big?

[prr]

Is there a sweet spot in terms of the size of the dictionary, beyond which, things slow down either LO or merely the context menu/spell check functionality?

Not trying to start any rumors---if this isn't the case, then please let me know. But I'm wondering if any longtime, hardcore users of AOO or LO have noticed that things slow down when their dictionary gets too big. I know to use the dictionary to autocorrect to names that I would commonly use or mistype. But I'm just wondering about the size of the dictionary, overall. If I do this every time I have a word that LO doesn't recognize, can this slow things down?

[Bill]

I assume that at some point, a large dictionary would slow things down, but it probably depends on the available memory and speed of the computer. I don't know if there is any documentation on this subject.

[prr]

Thanks. I wouldn't expect documentation on this, but I was wondering if some OO power users had encounted or noticed any slowdown, that's all.

[floris v]

I don't think it will, in a sorted dictionary you can search at logarithmic speed, so that the time taken would go up a little every time you double the size. That's because the search routine will first test against the word in the middle of the dictionary, then look in the middle of the remainder, and so on, until you are left with a search area with a single word in it. The number of rules for composing words from a base (compound nouns, rules for plural, conjugation of verbs, etc) and the number of words with such rules attached will probably have an effect.
Don't worry about this, just remember that it will always pay to proofread your work, even when the spell check doesn't produce anything.

[RoryOF]

I always place words for a specific project in a User dictionary, rather than add them to the main dictionary.

[John_Ha]

I very, very much doubt it.

The dictionary is arranged in alphabetical order so once you type the first character of the word, the dictionary effectively jumps to it, skipping over all the others. Similarly for the other characters in the word. This method of searching requires very little resource.

There are 59,000 lines in the Eng-GB dictionary, where the suffixes after the / define additional endings for the stem word.

Code: Select all

abject/PY
abjection/MS
abjectness/S
abjuration/M
abjure/nyRSDGN
ablate/SDG
ablation/M
ablaze
able-bodied
able/nVvYNT
abloom
ablution/SM
abnegate/NGnDS
abnegation/M
abnormal/Y
abnormality/SM
aboard
abode/MS
abolish/DGLRS
abolition/3MS
abolitionism/SM
abolitionist
abominable/Y
abominate/DnNSG
abomination/M
aboriginal
aboriginals/M
Aboriginals/M
aborigine/oSM
abort/DVGvSu

I did a test looking at CPU usage on my PC when I pasted "Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.", every word of which threw up a spelling error. The CPU usage increased from about 3% to about 8% for a second or so, which is negligible. I could see no difference when I pasted English text.

Similarly, the user defined dictionary is also arranged in alphabetical order so will have very little effect. The user dictionary does not use suffixes after the stem word - each word is defined fully.

If this solves the problem, please click the Edit button on your original post and add [Solved] in front of your subject.

[RoryOF]

The existing English dictionaries being very comprehensive, I doubt that the relatively small number of words a user might add will affect their overall speed significantly.

[John_Ha]

As an indication of why searching the dictionary is trivial compared with other things happening at the same time, consider what happens when you type a single character.

Each character is defined by a set of equations and the PC evaluates these equations and draws the lines on the outside of the characters. The PC then fills in all the insides. The PC then converts the picture so constructed to the actual pixels needed at the current zoom level and displays the character. In Windows, the edge pixels are then adjusted to minimise the "ragged edge" by changing their colour so as to use the individual colour pixel closest to the edge, and not the other two pixels. The eye does not see the edge colour and a better, less ragged edge is drawn.

I took this diagram from Parameterizable Fonts Based on Shape ... - Infoscience - EPFL.

[floris v]

I always place words for a specific project in a User dictionary, rather than add them to the main dictionary.

If you search in a large number of user dictionaries, all with rather few words, that might increase the search time.

Another things is, if you regularly use sets of words that differ only by one letter, but with totally different meanings, you might want to keep them out of your dictionaries, so that you will be forced to check you used the right one every time.

[prr]

Well thank you all for some very thorough answers!