Tryag File Manager
Home
-
Turbo Force
Current Path :
/
usr
/
lib
/
python2.4
/
site-packages
/
mx
/
BeeBase
/
Upload File :
New :
File
Dir
//usr/lib/python2.4/site-packages/mx/BeeBase/BeeDict.pyc
mò !N5?c @ s¸ d Z d k Z d k Z d k Z d k Z d k l Z e i Z d k Te i Z e i Z d Z d e i f d „ ƒ YZ d e f d „ ƒ YZ d e f d „ ƒ YZ d f d „ ƒ YZ d f d „ ƒ YZ d e f d „ ƒ YZ d e f d „ ƒ YZ e e ƒ d e f d „ ƒ YZ e e ƒ d e f d „ ƒ YZ e e ƒ d „ Z e d j o^ e d ƒ Z d e d <d e d <d e d <e d ƒ Z d e d <d e d <d e d <[ [ n d S(! s/ BeeDict - On-disk dictionary like databases N( s Tools( t *iè t Errorc B s t Z RS( N( t __name__t __module__( ( ( t6 /usr/lib/python2.4/site-packages/mx/BeeBase/BeeDict.pyR s t RecreateIndexErrorc B s t Z d Z RS( s… This error is raised in case the index for a dictionary was not found and/or needs to be recreated by running recovery. ( R R t __doc__( ( ( R R s t RecoverErrorc B s t Z d Z RS( sq This error is raised in case the storage for a dictionary was found to be in an inconsistent state. ( R R R ( ( ( R R " s t BeeBaseDictc B sg t Z d Z d Z e Z e Z d Z e Z d Z d Z e Z e Z d „ Z d „ Z d „ Z e d „ Z d „ Z d „ Z d „ Z d „ Z d d „ Z e e d! d „ Z d d „ Z d „ Z d" e e e d „ Z d „ Z e e e d „ Z e e e e d „ Z e d „ Z! d „ Z" d „ Z# d „ Z$ d „ Z% d „ Z& d „ Z' d „ Z( e d „ Z) d „ Z* RS(# s Base class that collects some method that all dicts can usually share without modification. The base class implements a transaction based caching mechanism. The cache stores data in the form key:(state,value) where state is: * 0 for read but not yet modified * 1 for modified * 2 for deleted * 3 for key read, value still on disk The .commit method has to implement the needed disk mechanism to write the cache data to disk storage. t NoNamei c C s h | _ d S( s Inits the cache. N( t selft cache( R ( ( R t __init__K s c C s t | i ƒ S( N( t lenR t index( R ( ( R t __len__Q s c C sg | i pY | i ƒ x? | i | i f D]+ } | d j o | i ƒ | i ƒ q' q' Wd | _ n d S( sŽ Flush buffers and close. This issues a .rollback(), so the current transaction is rolled back. i N( R t closedt rollbackR t storaget objt Nonet flusht close( R R ( ( R R U s c C s6 | i p( y | i ƒ Wq2 | j o q2 Xn d S( s) Make sure the object is closed. N( R R R t AttributeError( R R ( ( R t __del__e s c C sF | i d j o | i i ƒ n | i d j o | i i ƒ n d S( s Flush buffers to disk. N( R R R R R ( R ( ( R R r s c C s d | i i | i t | ƒ f S( Ns <%s instance for "%s" at 0x%x>( R t __class__R t namet id( R ( ( R t __repr__{ s c C s | i i ƒ | i ƒ d S( sÇ Commit all changes and start a new transaction. This method does not implement any disk operation, but takes care of managing the cache. Post-override. N( R R t clearR ( R ( ( R t commit s c C s | i i ƒ | i ƒ d S( sÇ Take back all changes and start a new transaction. Overriding is normally not needed. This method only takes care of managing the cache. Post-override. N( R R R R ( R ( ( R R s i i c C sM | i o? x< | i i ƒ D]' \ } \ } } | | j o d Sq q Wn d S( st Return true in case the current transaction includes changes to the database, false otherwise. i i N( R R t itemst keyt statet valuet modified( R R# R" R R! ( ( R t changed s i c C sÆ | i } x8 | i ƒ D]* \ } \ } } | | j o | | =q q W| | ƒ } | | j oe | i o | i ƒ d Sn t t d | ƒ | d | j o$ t t d ƒ | i ƒ t d ‚ q n d S( s˜ Remove all unmodified entries from the cache. If self.autocommit and the cache cleanup did not reduce the size below the MAXCACHESIZE limit, a .commit() will be executed automatically. Otherwise the processing continues. If the cache size hits 2 * MAXCACHESIZE, a forced .rollback() is done and an error raised. Ns* Cache overflow: %i modified items in cachei s$ Cache overflow: forced a .rollback()s, cache overflow; last transaction rolled back( R R R R R! R" t unmodifiedR t sizet MAXCACHESIZEt autocommitR t logt SYSTEM_ERRORR R ( R R R' R% R R! R" R R& ( ( R t free_cacheª s$ c C s t d ‚ d S( sù Read and return the value corresponding to key. If checkonly is true, only the availability of a value has to be checked. Override this method with an implementation that accesses the disk. s key not foundN( t KeyError( R R t checkonly( ( R t readË s c C s d | f | i | <d S( s¡ Save the item in the dictionary. Note that it is written to the cache first. Use .commit() to make sure it is written to disk. i N( R" R R R ( R R R" ( ( R t __setitem__× s c C s¦ y2 | i } | | \ } } | | j o | Sn Wn | j o n X| d j o | d ‚ n | i | ƒ } | | ƒ | j o | i ƒ n d | f | | <| S( s¤ Get an item from the dictionary. This first tries to read the item from cache and reverts to the disk storage if it is not found. i s key deletedi N( R R R R! R" t validstatesR, R. R R' R+ ( R R R0 R R, R' R R! R" ( ( R t __getitem__à s c C s; | i } | i | ƒ p | i | d ƒ n d | | <d S( s‹ Delete an item. The item is only marked for deletion. The next .commit() will make the action permanent. i i N( i N( R R t has_keyR R. R ( R R R ( ( R t __delitem__ÿ s c C s® y6 | i } | | \ } } | d j o d Sn d SWn | j o n Xy | i | ƒ } Wn | j o d Sn X| | ƒ | j o | i ƒ n d | f | | <d S( s Check if the dictionary has an item indexed by key. Successfully found items are put in the cache for fast subsequent access. i i i N( R R R R! R" R, R. R R' R+ ( R R R R, R' R R! R" ( ( R R2 s" c C s® y6 | i } | | \ } } | d j o | Sn | SWn | j o n Xy | i | ƒ } Wn | j o | Sn X| | ƒ | j o | i ƒ n d | f | | <| S( sß Get item indexed by key from the dictionary or default if no such item exists. This first tries to read the item from cache and reverts to the disk storage if it is not found. i i N( R R R R! R" t defaultR, R. R R' R+ ( R R R4 R R, R' R R! R" ( ( R t get* s" c C s[ | i o | i ƒ o t d ‚ n | i i | d ƒ } | d j o | Sn t | | ƒ S( sf Return a cursor instance for this kind of dictionary. In case the key is not found, default is returned instead. Note that cursors operate with the data on disk meaning that any uncommitted changes will not be seen by the cursor. Has to be overridden to return a suitable cursor object. s uncommitted data existsN( R R R$ R R t cursorR R R4 t BeeBaseDictCursor( R R R4 R6 ( ( R R6 K s c C s | i i ƒ \ } } } | | S( s¨ Determine the amount of garbage in bytes that has accumulated in the storage file. This amount would be freed if .collect() were run. N( R R t statisticst validt oldt invalid( R R9 R: R; ( ( R t garbage_ s c C sy | i o t d ‚ n | i o | i ƒ o t d ‚ n | i ƒ t t d | ƒ | i i | i ƒ | i i ƒ d S( s8 Run the storage garbage collector. Storage collection can only be done for writeable dictionaries and then only if the current transaction does not contain any pending changes. This can take a while depending on the size of the dictionary. s dict is read-onlys uncommitted data existss Collecting %sN( R t readonlyR R R$ R R) t SYSTEM_INFOR t collectt collect_callbackt end_transaction( R ( ( R R? i s c C s t d ‚ d S( sÛ Internal callback used to update the index when the storage garbage collector moves a record. This method must be overridden to account for the different indexing schemes. s# .collect_callback() not implementedN( R ( R t old_positiont new_positiont raw_data( ( R R@ s c C sA t t d | ƒ | i i ƒ | i i | i ƒ | i i ƒ d S( s; Recover all valid records and recreate the index. s Recovering %sN( R) R> R R R R t recovert recover_callbackRA ( R ( ( R RE Œ s c C s t d ‚ d S( sc Internal callback used to update the index when the storage recovery collector finds a record. old_position is only given as additional information, new_position contains the position of the found record. This method must be overridden to account for the different indexing schemes. s# .recover_callback() not implementedN( R ( R RB RC RD ( ( R RF ˜ s c C s1 t t d | ƒ | i i ƒ p t d ‚ n d S( s Checks the consistency of the index and raises an RecreateIndexError in case the index is found to be inconsistent. Validation takes places for the data on disk. The in memory data is not checked. s Validating index for %ss index is inconsistentN( R) R> R R t validateR ( R ( ( R t validate_index¦ s c C s t t d | ƒ d S( sg Checks the consistency of the storage and raises an RecoverError in case it is found to be inconsistent. Validation takes places for the data on disk. The in memory data is not checked. XXX This currently does nothing since storages don't support validation yet. s Validating storage for %sN( R) R> R ( R ( ( R t validate_storage´ s c C s | i ƒ | i i | ƒ d S( s¼ Issues a backup request to the storage using archive which defaults to the storage filename + '.backup'. This causes an implicit .rollback() to be done. N( R R R t backupt archive( R RK ( ( R RJ Ä s c C s t d ‚ d S( sg Restores the storage from an archive file. XXX Currently does NOT work. s .restore() not implemented !N( t SystemError( R RK ( ( R t restoreÏ s ( i i ( i i ( i i (+ R R R R R R R R R R= R( t FirstKeyt LastKeyR R R R R R R R R R$ R R' R+ R. R/ R, R1 R3 R2 R5 R6 R< R? R@ RE RF RH RI RJ RM ( ( ( R R + sH ! ! R7 c B sY t Z d Z e Z d „ Z e d „ Z d „ Z d „ Z d „ Z d „ Z d d „ Z RS( sþ Cursor for BeeBaseDict subclasses. The read/read_key/write methods assume that the dictionary uses a BeeKeyValueStorage instances as storage facility. Note that modifying the targetted dictionary while using a cursor can cause the cursor to skip new entries or fail due to deleted items. Especially deleting the key to which the cursor currently points can cause errors to be raised. In all other cases, the cursor will be repositioned. c C s. | | _ | | _ | i | _ | i | _ d S( N( t dictR R6 R R" ( R RP R6 ( ( R R ê s c C s | i i i | ƒ | _ } | od | i | j oT xQ | i ƒ o6 | i | j o Pn | i | j o t d ‚ q6 q6 Wt d ‚ n | | _ | | _ d S( sÔ Position the index cursor to index[key]. If value is given, index[key] == value is assured. key may also be FirstKey or LastKey (in which case value has to be None). s key not foundN( R RP R R6 R R" t nextR, ( R R R" R6 ( ( R t positionñ s c C s^ | i } | i p# | i | i | i ƒ | i } n | i ƒ } | i | _ | i | _ | S( s‘ Moves to the next entry in the dictionary. Returns true on success, false if the end-of-data has been reached. N( R R6 R9 RR R R" RQ t rc( R R6 RS ( ( R RQ s c C s^ | i } | i p# | i | i | i ƒ | i } n | i ƒ } | i | _ | i | _ | S( s• Moves to the previous entry in the dictionary. Returns true on success, false if the end-of-data has been reached. N( R R6 R9 RR R R" t prevRS ( R R6 RS ( ( R RT s c C s | i | i S( s# Reads the object from the dict to which the cursor currently points. This method assumes a BeeKeyValueStorage instance in self.storage and an index that holds copies of the key objects. Override if this is not the case. N( R RP R ( R ( ( R R. ' s c C s | i i i | i ƒ S( s¡ Reads the key object from the dict to which the cursor currently points. This method is intended for dictionaries that store hash values in the index. Normally, accessing the instance variable .key would give the current key object. This method assumes a BeeKeyValueStorage instance in self.storage. Override if this is not the case. N( R RP R t read_keyR" ( R ( ( R RU 3 s i c C s | | i | i <d S( sš Writes the object to the dict under the key to which the cursor currently points. The new data is not written to disk until the dictionaries current transaction is committed. This method assumes a BeeKeyValueStorage instance in self.storage and an index that holds copies of the key objects. Override if this is not the case. N( t objectR RP R ( R RV t minsize( ( R t writeA s ( R R R R R6 R RR RQ RT R. RU RX ( ( ( R R7 Ú s t BeeDictc B s• t Z d Z d d d d d e i d „ Z d „ Z e d „ Z e i d „ Z d d „ Z d „ Z d „ Z d „ Z e d „ Z d „ Z d „ Z RS( s On-disk dictionary that uses a "hash to address" index. Both Keys and values must be pickleable. Keys also have to be hashable. Both can have arbitrary size (keys shouldn't be too long though). Hash collisions are treated by sequential reads of all records with the same hash value and testing for equality of keys. This can be expensive ! i c C s | | _ t i | d d | d d d | d | d | ƒ| _ | o d } n% | i i p | o d } n d } y, t i | d d d d | d d ƒ| _ Wn t j o t d | ‚ n Xd | _ | | _ | | _ | o | i ƒ | i ƒ n | | ƒ d S( s¥ Create an instance using name as basename for the data and index files. min_recordsize is passed to the BeeStorage as indicator of the minimum size for data records. readonly can be set to true to open the files in read-only mode, preventing any disk modifications. To open the dictionary in recovery mode, pass a keyword recover=1. Then run .recover() and reopen using the normal settings. If autocommit is true the cache control will do an automatic .commit() whenever the transaction log overflows. If validate is true, the dictionary will run a validation check after having successfully opened storage and index. RecreateIndexError or RecoverError exceptions could be raised in case inconsistencies are found. s .datt lockR i t min_recordsizeR= RE i i i s .idxt dupkeyst filemodet sectorsizei s! Index for BeeDict "%s" is missingN( R R t BeeStoraget BeeKeyValueStorageR= R[ RE R R] t is_newt BeeIndext BeeIntegerIndexR t IOErrorR t collisionsR( RG RH RI t basemethod( R R R[ R= RE R( RG Rf R] ( ( R R _ s6 c C sX xQ | i ƒ oC | i | j o d Sn | | i i | i ƒ j o | i Sq q Wd S( s¬ Find the address of a key by scanning the hash value set pointed to by the cursor. Returns the address of the found key or None. N( R6 RQ R t hashvalueR R R RU R" ( R R6 Rg R ( ( R t find_address s c C s» | t j p | t j o | i i | | ƒ Sn t | ƒ } | i i | d ƒ } | d j o | Sn | i } | | i i | ƒ j o. | i | | | ƒ } | d j o | Sq· n | S( s¬ Return an index cursor pointing to key. In case the key is not found, default is returned. XXX Should use this method in more places... N( R RN RO R R R6 R4 t hashRg R R" t addressR RU Rh ( R R R4 Rg R6 Rj ( ( R t index_cursor² s c C s t o t t d | i ƒ n | i } | i } | i } | i } | i } | i } | i } | i } | i } x| i i ƒ D]ò\ } \ } } | d j ot o t i t d | | ƒ n | o t d ‚ n t | ƒ } | | d ƒ } | d j oš | i } | | | ƒ j o) | i d | _ | i | | | ƒ } n | | | | ƒ } | | j o q n | d j o | | | | ƒ q¯| | | <qq| | | ƒ } | | | <q | d j o± t o t t d | ƒ n | o t d ‚ n t | ƒ } | | ƒ } | i } | | | ƒ j o3 | i | | | ƒ } | d j o t! d ‚ qVn | | ƒ | | | ƒ q q W| i" ƒ | | ƒ d S( s9 Commit all changes and start a new transaction. s" Committing all changes for "%s"...i s writing key "%s":s dict is read-onlyi s deleting key "%s"s key not foundN($ t __debug__R) t SYSTEM_DEBUGR R R R6 Rk t updatet index_updatet deletet index_deleteR RU t storage_read_keyRX t storage_writet storage_deleteR= R R R R! R" RV R Ri Rg R Rj Re Rh t new_addressR, RA Rf ( R Rf Ru Rj Rk R Rg Rr R R= R! R Ro Rt Rs R" R6 Rq ( ( R R Ê sd c C sš t | ƒ } | i i | ƒ } | i } | | i i | ƒ j o3 | i | | | ƒ } | d j o t d ‚ qs n | o d Sn | i i | ƒ d Sd S( Ns key not foundi ( Ri R Rg R R R6 R" Rj R RU Rh R R, R- R. ( R R R- Rg R6 Rj ( ( R R. s c C sª g } | i i } xA | i i ƒ D]0 \ } \ } } | d j o | i | ƒ q" q" W| i i } xA | i i ƒ D]0 } | | ƒ } | | ƒ p | i | ƒ qr qr W| S( s™ Return a list of keys. The method does not load any data into the cache, but does take notice of uncommitted changes. i N( t lR R RU R R R R! R" t appendR2 t key_addedR t valuesRj ( R Rv R" Rx RU R! R Rj ( ( R t keys( s c C s° g } | i i } xA | i i ƒ D]0 \ } \ } } | d j o | i | ƒ q" q" W| i i } xG | i i ƒ D]6 } | | ƒ \ } } | | ƒ p | i | ƒ qr qr W| S( s› Return a list of values. The method does not load any data into the cache, but does take notice of uncommitted changes. i N( Rv R R R. R R R R! R" Rw R2 Rx R Ry Rj ( R R. Rv R" Rx R! R Rj ( ( R Ry > s c C s¼ g } | i i } xG | i i ƒ D]6 \ } \ } } | d j o | i | | f ƒ q" q" W| i i } xM | i i ƒ D]< } | | ƒ \ } } | | ƒ p | i | | f ƒ qx qx W| S( sš Return a list of items. The method does not load any data into the cache, but does take notice of uncommitted changes. i N( Rv R R R. R R R R! R" Rw R2 Rx R Ry Rj ( R R. Rv R" Rx R! R Rj ( ( R R T s c C sX | i o | i ƒ o t d ‚ n | i | d ƒ } | d j o | Sn t | | ƒ S( s+ Return a cursor instance for this kind of dictionary. Note that cursors operate with the data on disk. The method will raise an error in case there are uncommitted changes pending. In case the key is not found, default is returned instead. s uncommitted data existsN( R R R$ R Rk R R R6 R4 t BeeDictCursor( R R R4 R6 ( ( R R6 j s c C si t | i i | ƒ ƒ } y | i i | | | ƒ Wn0 t j o$ t t d | ƒ | | i | <n Xd S( ss Internal callback used to update the index when the storage garbage collector moves a record. sB Collect callback detected a missing key in index: %i; restoring itN( Ri R R t decode_keyRD R R Rn RC RB R, R) t SYSTEM_WARNING( R RB RC RD R ( ( R R@ } s c C s# | | i t | i i | ƒ ƒ <d S( s Internal callback used to update the index when the storage recovery collector finds a record. old_position is only given as additional information. new_position contains the position of the found record. N( RC R R Ri R R| RD ( R RB RC RD ( ( R RF Œ s ( R R R R R Rh R Rk R R. Rz Ry R R6 R@ RF ( ( ( R RY R s A L R{ c B s, t Z d Z d „ Z d „ Z d d „ Z RS( sØ Cursor for BeeDict instances. Since the keys are not kept directly in the index, .read_key() must be used to obtain the real key objects instead of querying the instance variable .key. c C s# | i i i | i ƒ } | i | S( sZ Reads the object from the dict to which the cursor currently points. N( R RP R RU R" t keyobj( R R~ ( ( R R. ¢ s c C s | i i i | i ƒ S( s^ Reads the key object from the dict to which the cursor currently points. N( R RP R RU R" ( R ( ( R RU ª s i c C s) | i i i | i ƒ } | | i | <d S( sÝ Writes the object to the dict under the key to which the cursor currently points. The new data is not written to disk until the dictionary's current transaction is committed. N( R RP R RU R" R~ RV ( R RV RW R~ ( ( R RX ± s ( R R R R. RU RX ( ( ( R R{ ™ s t BeeStringDictc B sƒ t Z d Z d d d d d d e i d „ Z e i d „ Z d d „ Z d „ Z d „ Z d „ Z e d „ Z d „ Z d „ Z RS( s§ A dictionary that is indexed by limited size strings. Though the keys must be strings of limited size, the values can be any pickleable object. i i c C s | | _ t i | d d | d d d | d | d | ƒ| _ | o d } n% | i i p | o d } n d } y2 t i | d d d d | d | d d ƒ| _ Wn t j o t d | ‚ n X| | _ | | _ | o | i ƒ | i ƒ n | | ƒ d S( sÿ Create an instance using name as dictionary name. Two files will be created: name.dat and name.idx. keysize gives the maximal size of the strings used as index keys. min_recordsize gives a hint to the expected typical size of (key,value) pickles: all records will have at least this size. To open the dictionary in recovery mode, pass a keyword recover=1. Then run .recover() and reopen using the normal settings. If autocommit is true the cache control will do an automatic .commit() whenever the transaction log overflows. If validate is true, the dictionary will run a validation check after having successfully opened storage and index. RecreateIndexError or RecoverError exceptions could be raised in case inconsistencies are found. XXX Save keysize in storage file header. s .datRZ R i R[ R= RE i i i s .idxR\ t keysizeR] R^ i s0 Index for BeeStringDict "%s" could not be openedN( R R R_ R` R= R[ RE R R] Ra Rb t BeeStringIndexR€ R Rd R R( RG RH RI Rf ( R R R€ R[ R= RE R( RG Rf R] ( ( R R È s6 c C s¨ t o t t d | i ƒ n | i } | i } | i } | i } | i } | i } x9| i i ƒ D](\ } \ } } | d j o t o t i t d | | ƒ n | | d ƒ } | d j o4 | | | | ƒ } | | j o qd n | | | <qŒ| | | ƒ } | | | <qd | d j om t o t t d | ƒ n | o t d ‚ n | | d ƒ } | d j o | | ƒ | | =qŒt d ‚ qd qd W| i ƒ | | ƒ d S( s9 Commit all changes and start a new transaction. s" Committing all changes for "%s"...i s writing key "%s":i s deleting key "%s"s dict is read-onlys key not foundN( Rl R) Rm R R R R5 t index_getR RX Rs Rp Rt R= R R R R! R" RV R Rj t new_addrR R, RA Rf ( R Rf R‚ R Rs Rj R R! R" R= Rƒ R Rt ( ( R R sD c C s4 | i | } | o d Sn | i i | ƒ d Sd S( Ni ( R R R Rj R- R R. ( R R R- Rj ( ( R R. A s c C s1 | i o | i ƒ o t d ‚ n | i i ƒ S( sÉ Return a list of keys. The method will raise an error if there are uncommitted changes pending. Output is sorted ascending according to keys. s uncommitted data existsN( R R R$ R R Rz ( R ( ( R Rz I s c C sk | i o | i ƒ o t d ‚ n g } | i i } x. | i i ƒ D] } | i | | ƒ d ƒ qF W| S( sË Return a list of values. The method will raise an error if there are uncommitted changes pending. Output is sorted ascending according to keys. s uncommitted data existsi N( R R R$ R Rv R R. R Ry Rj Rw ( R R. Rj Rv ( ( R Ry V s c C sg | i o | i ƒ o t d ‚ n g } | i i } x* | i i ƒ D] } | i | | ƒ ƒ qF W| S( sÊ Return a list of items. The method will raise an error if there are uncommitted changes pending. Output is sorted ascending according to keys. s uncommitted data existsN( R R R$ R Rv R R. R Ry Rj Rw ( R R. Rj Rv ( ( R R g s c C s[ | i o | i ƒ o t d ‚ n | i i | d ƒ } | d j o | Sn t | | ƒ S( s! Return a cursor instance for this kind of dictionary. Note that cursors operate with the data on disk meaning that any uncommitted changes will not be seen by the cursor. In case the key is not found, default is returned instead. s uncommitted data existsN( R R R$ R R R6 R R R4 t BeeStringDictCursor( R R R4 R6 ( ( R R6 x s c C sc | i i | ƒ } y | i i | | | ƒ Wn0 t j o$ t t d | ƒ | | i | <n Xd S( ss Internal callback used to update the index when the storage garbage collector moves a record. sB Collect callback detected a missing key in index: %i; restoring itN( R R R| RD R R Rn RC RB R, R) R} ( R RB RC RD R ( ( R R@ ‹ s c C s | | i | i i | ƒ <d S( sm Internal callback used to update the index when the storage recovery collector finds a record. old_position is only given as additional information. new_position contains the position of the found record. XXX What if the recovery collector find two records with the same key ? N( RC R R R R| RD ( R RB RC RD ( ( R RF š s ( R R R R R R R. Rz Ry R R R6 R@ RF ( ( ( R R À s !A8 R„ c B s t Z d Z d „ Z RS( s) Cursor for BeeStringDict instances. c C s | i S( s^ Reads the key object from the dict to which the cursor currently points. N( R R ( R ( ( R RU ° s ( R R R RU ( ( ( R R„ ¬ s c O sÞ y t | | | ƒ } | SWn: t i j o n& t j o n t j o n Xt t d | d ƒ | i ƒ } d | d <d | d <d | d <t | | | ƒ } | i ƒ ~ t | | | ƒ } t t d | d ƒ | S( s« Wrapper that runs recovery whenever needed. This can still fail, but at least many "normal" failures can be handled this way automatically. s, Database "%s" is corrupt; trying recovery...i R= RG i RE s( Successfully recovered the database "%s"N( t applyt Classt argst kwst dR_ t RunRecoveryErrorR R R) R* t copyt new_kwsRE R} ( R† R‡ Rˆ R‰ RŒ ( ( R t AutoRecover» s. t __main__s test-BeeStringDictt Svetat Marct Petrat Thorstent Leesat Christians test-BeeDict( R t exceptionst UserDictRb R_ t mxt Toolst freezet mx.LogRN RO R' t StandardErrorR R R R R7 RY R{ R R„ R R t sR‰ ( R™ R7 RO R_ R R{ R' R R– RN R R Rb RY R‰ R R Rœ R„ R• R˜ ( ( R t ? sB ÿ °xÿ H# è %