Ver. 2026.3.0.0 - SmartMapper extensions, multi site support and tons of other improvements

New in version 2026.3.0.0 (6th of November, 2026)

SmartMapper

• Clearer last page of the SmartMapper wizard. The page Search & replace keys created now starts with a summary of the result: the number of Search & Replace keys and of documents with keys of the run (and the coverage, if the keys were created in this session). The choice of the documents to process is listed one option under the other, each with a short description, and Only documents with SmartMapper keys is marked as the fastest choice. It works on a scan of the source as well as on a scan of the target location.
• Additional runs for ID-based links. SmartMapper can now resolve ID-based links whose target document was scanned in a different run. This covers sharing links, Document ID links (DocIdRedir.aspx?ID=) and sourcedoc links. Previously these links were resolved only if the target was scanned in the same run; otherwise they were reported as unresolved (e.g. SharingTokenUnresolved or IDMatchNotFound). This mainly helps migrations where each site is scanned as its own run.
  • New wizard step "Additional runs for ID-based links resolution". It appears after the source run selection when the selected run was scanned on SharePoint. Select the other SharePoint scanning runs that SmartMapper should also search when resolving ID-based links. The keys, mapping rules and results still belong to the selected run.
  • Multi-select and Select all. Select several runs with Ctrl/Shift+click, or use Select all / Clear selection.
  • Optional. If no additional runs are selected, SmartMapper works exactly as before.
  • Remembered selection. Your selection is kept when you move back and forth in the wizard. If you choose a different source run, the selection is cleared.
  • Command-line / silent runs. The Command Line Parameters form now has an Additional runs list with a Check All option, instead of a free-text field. The selected runs are saved to the XML configuration, e.g. <SMARTMAPPERADDITIONALRUNS>6,7,8</SMARTMAPPERADDITIONALRUNS>.
  • Traceability. For each resolved ID-based link, the Resolved Links view and export show which run the target document was found in (column IdSourceRunID).
  • Processing order: additional runs are processed from the latest to the oldest.
• Sharing links are decoded. Current SharePoint sharing links (e.g. https://tenant.sharepoint.com/:w:/s/site/IQ...) contain the ID of the target document. SmartMapper now reads that ID, so these links can be resolved even when the scan did not capture the sharing link itself. Such links are shown with the resolution method ShareTokenDecode. Sharing links in the older format (the token after the site name starts with "E") are still resolved only when the scan captured that sharing link.
• Extended support for UniqueID and GUID based links.
• Links shared with "Can review" and "Can't download" are converted; Unresolved Links show the file name of the document containing the link.
• Links relative to the root of a drive or share are converted. Excel and Word store a link to a file on the same drive as, for example, \Finance\Reports\Budget.xlsx, without the drive letter. SmartMapper now resolves such links against the drive or network share of the document that contains them and converts them with your migration rules. Previously they were reported as RootRelativeUnresolved and left unchanged. This is done only when the target was part of the scan, so links that really point to the root of a web site are still left alone. On a customer scan of 1.1 million documents this converted 44,491 additional links.
• Links that start with the scanned folder are changed even if their target no longer exists. A link to a file on the same drive is often saved without the drive letter (for example \Finance\Reports\Old folder\Book.xlsx). SmartMapper changed such a link only when its folder was found in the scan. Now it is also changed when the link starts with the folder you scanned (for example \Finance when you scanned Z:\Finance), exactly like the same link written with the drive letter (Z:\Finance\...). The file it points to was usually already missing before the migration, so the new link points to the place where it would be after the migration. Links that do not start with the scanned folder (web pages, other folders of the drive) are not changed, as before.
• SmartMapper keys of type Relative no longer change part of another link. A Relative key (for example Picard/Book.xlsx) was also applied inside a longer link of the same document that ends with the same text (for example /Finance/March/Picard/Book.xlsx), which produced a link with two paths glued together. Relative keys are now applied only to a link that starts with them, in the same way as same-folder links. Both kinds of keys are also no longer applied after "@" or ":" (for example inside mailto: addresses or C:\... paths) or to a file name that follows a folder path in the same link (for example a pivot table source read from SharePoint). Absolute keys are not affected.
• "Formulas (Excel)" is no longer used for a SmartMapper replacement. With Formulas (Excel) selected in the Search & Replace options, the SmartMapper keys were also applied to cell formulas. A formula that uses another workbook already shows the changed link source, so the same key could be applied to it a second time and Excel got an additional link with the path twice (for example https://.../Folder/https://.../Folder/Book.xls). While SmartMapper keys are used, Formulas (Excel) can no longer be selected; if it was already selected, ReplaceMagic de-selects it after the SmartMapper wizard and shows a message - all before the replacement starts. In a command-line (silent) run the option is switched off right after the XML file is read and this is written to the log (the XML file is not changed). Links to other workbooks are changed by Link Sources and the formulas follow them. Search & Replace without SmartMapper keys is not affected.
• SmartMapper checks your migration rules before it creates the keys. The log now contains a "Mapping rules check" section. It shows where each scanned folder ends up in the target, and warns about rules that are not used, rules that send links outside the scanned folders or local paths such as C:\Users to SharePoint, and other drives or shares that point to the same folders as the scanned one. If the database also contains a scan of the migrated location (an Ultra fast scan is enough), SmartMapper checks every rule against it, proposes the right rule when the files are not where a rule sends them (for example when only the contents of a folder were migrated), and finds folders the migration tool renamed (for example ~Rec Logs to _Rec Logs). Rules backed by such evidence are added automatically; your existing rules are never changed, and the previous rules are listed in the log.
• New wizard step "Suggested mapping rules". After the keys are created, the SmartMapper wizard lists the rules it added, with the reason for each. Accept them, or delete the ones you do not want and recreate the keys with one click. The step is skipped when no rule was suggested. The migration rules list shows the origin of every rule: User, Suggested or Accepted.
• Rules to check. After the keys are created, the page Suggested mapping rules also lists your existing mapping rules that map no scanned document and are probably wrong or not needed - rules for local paths (C:\Users, C:\Program Files ...), rules whose links all point outside the scanned folders, rules superseded by more specific ones, and unused rules - with the number of links and of files that change only because of each rule. Delete the wrong ones and recreate the keys with one click; nothing is deleted automatically.
• A rule is no longer listed under "Rules to check" when the scan is of the migrated location. If you scanned the documents where they are now (for example on SharePoint after the migration) and their links still point to the old location (for example C:\Work\Migration\...), every link of the mapping rule pointed outside the scanned folders. The rule was therefore listed as outside scan (probably wrong) and the log showed a warning - even though it was the rule that all the keys were made with. Now a link counts as verified when the rule maps it to a document of the same scan, so such a rule is no longer listed and the log shows how many of its links were verified this way.
• SmartMapper warns about special characters in folder names. When keys point to folders whose names contain characters that migrations usually replace (for example ~Rec Logs → _Rec Logs) and the scan of the migrated location does not contain those folders, SmartMapper shows a warning with the number of keys and files and recommends an Ultra fast scan of the migrated location. On the Create keys page you can say whether the migration replaced such characters (I don't know / Yes / No).
• The Resolved Links export shows whether the new link's file exists on SharePoint. If the database holds a ReplaceMagic scan (Scan Documents) of the SharePoint site, the export has a new column TargetInScan: Yes, No, Not checked (a file type or folder that the scan does not contain) or No scan of this site. The message at the end of the export shows how many new links were found and not found. A link whose file is not on SharePoint usually was already broken before the migration. Tip: scan the SharePoint site with all the file types your links point to (for example also PDF and Word), otherwise these links are only "Not checked".
• Warning before SmartMapper changes documents at their source location. When a SmartMapper replacement is started on documents that are still at their old (source) location - for example after importing a saved run of a file share scan - ReplaceMagic now asks first. The message Documents at the SOURCE location shows how many of the documents to process have SmartMapper keys and are at the source (for example 8 of the 17 documents to process have SmartMapper keys and are at their SOURCE location), an example with its new location, and whether the Backup option is on. No (the default) cancels the replacement, Yes changes the documents in place. This is correct only when you repair the links before the migration; after the migration, scan the target location and run the replacement there. Command-line (silent) runs do not ask: they write a warning to the log and continue.
• Automatic database check in SmartMapper. When you start the SmartMapper wizard, and again before keys are created, SmartMapper checks the database structure and updates it if needed. This also covers databases that are marked as up to date but are missing required columns, for example a database connected through the DB Wizard. If the update is not possible, for example because of missing permissions or a read-only database file, SmartMapper shows a clear message and does not start key creation.

Process only the documents that can change

• Import only the documents that can change. When you import a saved run (Saved RUN(s) → Import), a new row above the runs lets you choose: All documents (default), Only documents with links (documents in which the scan found hyperlinks, OLE links, link sources, pivot tables or Power Queries) or Only documents with SmartMapper keys (exactly the documents the replacement will change). You can scan everything, create the keys for everything, and still open only the documents that will change - for example 27,067 of 1.1 million scanned documents. The documents grid shows only the imported documents and a message tells you how many were imported. If the run has no link counts (Ultra fast scan) or no SmartMapper keys yet, all documents are imported. In command-line (silent) runs the same choice is in Silent run params → SmartMapper options → Documents, saved as <SMARTMAPPERDOCUMENTS>All|WithLinks|WithKeys</SMARTMAPPERDOCUMENTS>.
• Process only the documents that can change - also at the target. The last page of the SmartMapper wizard now has the choice Documents to process: All documents, Only documents with links or Only documents with SmartMapper keys. It applies to the replacement you start from the wizard, also to a scan of the target location after the migration: with the keys of the source run, only the migrated documents that will change are opened - the target does not have to be saved in the database. Before the replacement starts, ReplaceMagic shows how many documents will be processed and lets you choose between only those and all documents. Only documents with links now uses the link counts of the documents being processed (a full scan of the target works too), also in command-line (silent) runs.
  
  

Several SharePoint sites in one run (Multi Site extension)

• Several SharePoint sites in one run. In the SharePoint Selection form you can now add more than one SharePoint site (site collection) - a new site is added to the sites already in the list instead of replacing them. Each site keeps its own connection: its own user, authentication mode, Microsoft Information Protection settings and, if used, its own additional SharePoint users. Sites can be in different tenants. Select locations from all sites and scan and replace them in one run; every document is opened and saved through the connection of the site it belongs to.
• Remove a single site. Right-click a site in the SharePoint Selection form and choose Delete to remove only that site, together with its selected locations. The other sites stay as they are, and the sites that were already in the list keep their expanded or collapsed structure when you add a new site.
• Time zone per site. The SharePoint time zone is read for every site, so the original Modified date is restored correctly for documents of each site, also in command-line (silent) runs.
• Command-line (silent) runs with several sites. The XML configuration contains the main site as before and one <SHAREPOINTCONNECTION> block for every additional site. If a site cannot be connected, only its locations are skipped (written to the log). Existing configuration files work without changes.
• Link check across sites. Links that point to another site in the list are checked through the connection of that site.
• Import of saved runs with several sites. When documents of a saved run are imported and the original Modified By is kept, the user is now looked up on the document's own site. Before, documents of an additional site could get the wrong user.
  
  

SharePoint Link items (.url)

• SharePoint "Link" items and Internet Shortcuts (.url) are supported. A Link item in a SharePoint library (Create or upload > Link) is a small .url file that points to another document or page. Until now ReplaceMagic did not process these files, so after a migration the Link items still pointed to the old location.
  • Scan: the target of every .url file is shown in the Hyperlinks tab. LinkTitle is the name of the item exactly as in the Name column of the library (for example Document Control.url). The link is checked and used by SmartMapper like every other link.
  • Replace: switch on Target Path and/or SharePoint .url files on the LNK/URL page. Only the target of the link is changed - the rest of the file (icon and other settings) stays as it is. The option Update text does not change .url files. If a replacement would leave a target that is not a complete link (for example an empty text or a path without a server or drive), the file is not changed and the reason is written to the log.
  • SharePoint: when you click a Link item, SharePoint does not read the file - it uses an address it keeps for the item. ReplaceMagic now updates this address together with the file, so the Link opens the new location. The log shows .url: Link target in SharePoint updated for every changed item. In a library where ReplaceMagic does not check documents out, this second write makes one extra version of the item.
  • File server: .url files in folders are changed in the same way.
  • Windows Shortcuts (.lnk) on SharePoint are now skipped (listed among the skipped documents): they can be changed only in folders. Before, they were read as text and a key could change the binary file.
  • url files are counted as Shortcuts in the scan and replacement dashboards, and are available in all ReplaceMagic editions that support Windows Shortcuts (.lnk).
    
    

SharePoint processing, sign-in and throttling

• Improved processing of documents on SharePoint (metadata, try...catch messages with better description...).
• New parameter Save without extra version under Configuration → SharePoint 1 - save documents with one new version instead of two. Without this option, restoring the original metadata (modified date, editor, approval status) creates an additional version after each save. If a server does not support it (for example SharePoint 2013), only that server falls back to two versions; other sites in the same run, and later runs, are not affected.
• Automatic re-login and clearer error reporting.
  • Automatic re-login: if SharePoint stops accepting the sign-in during a run (HTTP 401/403), ReplaceMagic now pauses processing, signs in again and continues where it stopped. With Web Logon the sign-in window opens again and usually closes by itself; if SharePoint asks for a password or MFA, the user signs in. With user name/password and with app-only certificate authentication this happens without any user action.
  • No more wasted retries: a document that SharePoint refuses is skipped immediately, instead of being retried five times with a sign-in that no longer works.
  • Faster finish when sign-in cannot be restored: the remaining documents are skipped at once with a clear reason, so the run ends quickly and the documents can be processed again later.
  • Fixed: with app-only certificate authentication every request failed after about one hour, because the access token was never renewed. It is now renewed automatically.
  • Clearer log and skipped-documents list: they now state the real reason when a document is skipped because of sign-in problems, and what to check (for example, whether a document may still be checked out). The log also records why SharePoint refused the sign-in, which helps support find the cause.
  • New settings on the SharePoint tab of the configuration: how long the Web Logon window waits for the user, the pause between attempts, and the maximum number of attempts (defaults: 5 minutes, 5 minutes, 6 attempts).
• Throttled documents are no longer skipped at once. When SharePoint throttles a request (HTTP 429 or 503), ReplaceMagic now waits as long as SharePoint asks (the Retry-After value) and then repeats the same step for the same document, up to 3 times. Only if SharePoint still throttles is the document skipped, as before. Before, the throttled document was skipped immediately and only the following documents waited. The wait time is now also taken from SharePoint when a document is downloaded or a link is checked (before: always 120 seconds for HTTP 503), and the countdown in the status bar now starts at the full wait time.
• A document could be skipped after throttling while its sharing links were read. If SharePoint throttled the request that reads the sharing links of a document, the following download of the same document failed five times with Cannot invoke HTTP DAV request. There is a pending query. and the document was skipped. Now the sharing links are read again after the wait requested by SharePoint (up to 3 times), and the download is no longer blocked.
• Decrease workload during SharePoint throttling: one step per throttling wave. When SharePoint throttled several documents at the same moment (HTTP 429 or 503), every one of them halved the number of documents processed in parallel - for example from 16 straight down to 1, so the rest of the run was much slower than needed. Now the parallel processing is halved once per throttling wave: within the wait time that SharePoint requested when it was halved (Retry-After, at least 5 seconds) it is not reduced again. If SharePoint is still throttling after that time, it is halved again. The log shows WORKLOAD REDUCED (from N threads to M threads) once per wave. This applies when Decrease workload during SharePoint throttling? is selected in the configuration - now also during a replacement: before, only a scan reduced the parallel processing, and a replacement kept running with the full number of documents while SharePoint throttled.
• Waiting for SharePoint no longer counts as processing time. When SharePoint throttles a document, ReplaceMagic waits as long as SharePoint asks (Retry-After, up to 3 times) and then continues with the same document. This waiting time was counted in the document processing timeout, so a document that had to wait twice (for example 95 + 177 seconds) was stopped and listed among the skipped documents, although it had not been changed yet. Now the waiting time is not counted. If a document is still stopped, the log shows how long it waited (throttling wait not counted: N s).
• Default wait for SharePoint HTTP 503 is 99 seconds, and the log shows where the wait time comes from. When SharePoint answers HTTP 503 without a Retry-After value, ReplaceMagic now waits 99 seconds (before: 120 seconds). The log line Retry-After now says whether the time came from SharePoint (Retry-After: 120 (from SharePoint)) or is the ReplaceMagic default (Retry-After: 99 (SharePoint sent no Retry-After - ReplaceMagic default)). Before, both looked the same. The default of Throttling retry-after period in the configuration is now also 99 seconds; a value you have already saved stays as it is.
• "Keep last modified date" moved some dates by one hour. On sites in a time zone with daylight saving time, a document whose Modified date was in the other half of the year (for example a February date while the site was converted with summer time) got a Modified date one hour later after the replacement. ReplaceMagic now converts every date with the daylight saving rule of the site's time zone for that date. Dates that were already moved are not corrected automatically.
• App-Only authentication cannot be combined with Microsoft Information Protection. In the SharePoint Selection form, Use MIP is now switched off and locked when App-Only client secret or App-Only certificate is selected (also for saved connections). SharePoint does not accept saving documents with an encrypting sensitivity label from an App-Only connection, so such documents could be read but never saved. A command-line (silent) run with App-Only and MIP in the XML file writes a note to the log and runs without MIP (no sign-in window). To process protected documents, use Web Logon (modern authentication) with MIP.
• Extended MIP logging.
• Search & Replace in SharePoint menus fixed. Replacements in the SharePoint Quick Launch and Top Navigation Bar are now saved. Before, ReplaceMagic found the menu links during the scan, but the changes to the menu entries could not be saved in SharePoint. Sub-menu entries are now processed with the search options you selected (Whole word only, Match case, Wildcards, Regular expression) - before, these options were passed on wrongly for sub-menus. With several Search & Replace keys, menus and list items now use the right search and replace pairs (before, from the second key on, wrong pairs were used).
• SharePoint lists. Update SharePoint List Items hyperlinks works again: the replacement in list items stopped with an error ("Conversion from string ... to type 'Integer'"), and after a timeout the list was reported as skipped - no list item was changed. If list processing fails, the reason is now written to the log and to the skipped documents at once. Several keys that change the description of a hyperlink column are now all applied (before only the last one); if SharePoint refuses to save a page of list items, the log and the problems list say so and these changes are no longer reported as done; the Document ID link of a list item (_dlc_DocIdUrl) is never changed; the scan export shows the correct address of a list or library, also when the whole list is selected (for example https://tenant.sharepoint.com/Lists/Links; before https://Lists/tenant.sharepoint.com or only the site address); for attachments of list items the log no longer shows "Problem to read SharePoint Shared Settings" (attachments have no sharing links). At the end of a replacement the log now shows how many list hyperlinks were changed (SharePoint list hyperlinks changed: N).
• Correct addresses for list item attachments on sites and subsites that are not the root site of the tenant (the site path was repeated in the address).
• Sharing links created with "Can review" and "Can't download" are read (new columns ReviewLink and BlockDownloadLink in scan results, exports, CSV import, database and saved runs). Links for "Anyone" and "People in your organization" created in the new SharePoint Share dialog are now shown as well.
• Reading of SharePoint sharing links is faster: documents that inherit permissions need one request instead of two.
• Selecting several locations at once fixed. In the SharePoint Selection form, when you selected several nodes (Ctrl+click) and one of them was already in the list of selected locations, the following nodes were silently skipped. Now every selected node is added.
• SharePoint pages: the page structure is protected. When changing Wiki pages, modern (Canvas) pages and Summary Links, a short search&replace key could also change the hidden structure of the page - for example the key Old changed the internal attribute data-sp-controldata, and the modern page was shown empty and could not be edited. Now the keys change only the visible text and the links - also the links in web part settings (for example the links of a Quick Links web part). Names of tags, attributes and settings, internal values of web parts (for example the layout of a News web part), layout data and IDs are never changed. Texts stored only in web part settings, such as the title of a News web part, are not changed either. If a key with wildcards or a regular expression would still damage the structure, the page is not saved and the problem is listed in the Search & Replace results.
• SharePoint pages: \ to / and / to \ change only links. On Wiki pages, modern pages and Summary Links the Convert options now change only link addresses (for example href and src of links and pictures, and links in web part settings). Normal text on the page, for example a path like C:\Temp\Data written as text, keeps its backslashes. Before, the conversion was applied to the whole text of the page when another selected area (for example Summary Links) allowed the Convert options.
• Quick Launch and Top Navigation Bar: link check. With Check if hyperlink exists (or broken link check) the column HyperlinkExists of menu items showed a wrong result. Menu links are now checked with their full address; a menu item without a link shows Link is without address.
• Quick Launch and Top Navigation Bar: every site is processed once. When several scans were made in the same ReplaceMagic session, the menus of a site were processed once for every scan - the log showed the same site several times, and a key that is part of its own replacement (for example A > AB) could be applied more than once. Now the menus of each site are processed exactly once per replacement, also when a parent site with subsites and one of its subsites are both selected.
• SharePoint sign-in lost during a long run: the run waits for you instead of skipping documents.
  • If SharePoint stops accepting the sign-in and the automatic re-login does not succeed (for example, nobody answered the Web Logon window), ReplaceMagic now pauses the run and asks: Yes = sign in again, No = stop the run. While the question is open no document is skipped. Before, all remaining documents of the run were skipped, which on a large run could take hours and produce a very long list of skipped documents.
  • If the sign-in cannot be restored (you choose to stop, or in a command-line/silent run), the run is stopped. Documents that were already open are listed as skipped; run it again after signing in - documents that were already changed no longer contain the old links.
  • While ReplaceMagic signs in again, no new documents are started, and the waiting time no longer counts towards the document processing timeout (before, documents waiting for the sign-in were stopped as "long running").
  • Long runs (several days) with Web Logon can now sign in again as often as needed: the limit of three automatic re-logins applies only within one hour.
  • Recommendation: never run silent run if there is a risk that user credentials expire and re-logon is required.
• Shorter wait for a short SharePoint outage (HTTP 503). When SharePoint answers "503 Service Unavailable" without saying how long to wait (no Retry-After), ReplaceMagic now waits 10 seconds and tries the same document again; if such answers come again, it waits 20, 40 and 80 seconds. Only from the second such answer is the number of parallel documents reduced and processing paused. Before, every such answer meant 99 seconds of waiting and half of the parallel documents. When SharePoint does say how long to wait, ReplaceMagic waits exactly that long, also when SharePoint gives a date and time instead of seconds. The log shows the value SharePoint sent ("Retry-After (as sent by SharePoint)") and the request id.

SharePoint processing, sign-in and throttling

• Documents with an encoded character in the file name are processed correctly. A SharePoint document whose name contains a percent sign followed by two characters, for example "Report%20Final.docx" or "Link%25Sources.xlsx", was skipped with "(404) Not Found" in every scan when you sign in with a method other than App-Only. SharePoint read "%20" as a space and looked for a different document ("Report Final.docx"). If such a document existed in the same folder, ReplaceMagic read that other document instead, and a replacement could have saved the changed content into it. ReplaceMagic now sends the exact document name, so these documents are scanned and changed like any other document, and the other document is not touched. App-Only sign-in was not affected.
• A document that no longer exists on SharePoint is skipped at once. When SharePoint answers "(404) Not Found" for a document, for example because it was deleted or renamed after the scan, ReplaceMagic now skips it right away instead of trying 5 times with growing pauses. The log says "document not found on SharePoint (HTTP 404) - not retried" instead of "5th retry, download document failed".
• The log shows who decided how long to wait during SharePoint throttling. The line "Wait:" in the block IMPORTANT SHAREPOINT HEADERS and the line "SharePoint throttling (503): waiting ..." now name the source of the wait: "Retry-After from SharePoint" when SharePoint sent the time, "ReplaceMagic wait 1 of 4" when SharePoint sent no time (10, 20, 40, 80 seconds), or "configured Throttling retry-after period" when "Use SharePoint retry-after recommendation" is switched off. Before, the log always said "Retry-After", also when ReplaceMagic chose the time. How long ReplaceMagic waits did not change.
  
  

Link check

• Links to SharePoint sites you cannot access are no longer reported as broken. When a link points to a SharePoint site that the signed-in user cannot open (HTTP 401 or 403) - for example a site that is not in the SharePoint Selection form or a site in another tenant - the result is now Link not verified (code: 401, access denied) (or code 403) instead of Broken link, and the link is not counted as broken. Links to other web sites are checked as before. To check such links, add their site in the SharePoint Selection form.
• No more links without a result. If the SharePoint link check cannot be finished (for example when SharePoint throttles the requests), the result is now Link not verified (SharePoint check failed: ...) with the reason. Before, such links were shown without any result.
• Correct number of broken links per document. The number of broken links per document now matches the total in the result summary. Before, links checked over HTTP with the result Link OK (code: 200, OK) were counted as broken in the per-document number, and some failed checks were counted only per document and not in the total.
• More reliable link check on SharePoint. When documents are processed in parallel, every thread now checks SharePoint links through its own connection. Before, existing links could occasionally be reported as broken.
  
  

SharePoint illegal characters

• New option "Use optional illegal characters?" (Configuration → SharePoint 2): the characters # % & { } ~ are replaced only when this option is selected, while " * < > | ? are always replaced (when Replace SharePoint illegal characters? is on).
• Illegal characters are replaced only in the path of a link. Like the migration tools, ReplaceMagic now changes only the folder and file names of a link. The server address and everything after the file name stay as they are: the sheet name, cell range or table reference ('...\Stock.xlsx'!Data[#Data], [Plan.xlsx]Sheet #2), a bookmark or anchor (Report.xlsx#Sheet1!A1, Home.aspx#section2) and the query after "?". Encoded characters such as %20 stay as they are, in full links (https://company.sharepoint.com/Shared%20Documents/...) as well as in links without https:// (/Shared%20Documents/..., relative links). An encoded illegal character is replaced like the character itself (%23 = "#"), because it is part of the real file or folder name. In formulas and field codes only the quoted paths are converted. E-mail addresses (mailto:) are not changed. Tip: if your migration tool replaced only "#" and "%" (for example ShareGate), keep only these two in the optional list.
• Power Queries keep their formula when SharePoint illegal characters are replaced. With Replace SharePoint illegal characters and Power Queries selected (Configuration > SharePoint 2), the whole Power Query formula was converted, so its quotation marks became "_" and the query no longer worked. Now only the folder and file names in the paths of the formula are converted; the rest of the formula (quotation marks, https://, %20, delimiters) stays as it is.
  
  

Search & Replace options

• New option "Keys: Space<->%20" (first named "Keys: space = %20"). A link can be written with a space (Shared Documents) or with %20 (Shared%20Documents), and a search&replace key only finds the form it is written in. With the new option on the Search & Replace form, every key that contains a space is also used in its %20 form, and every key with %20 also in its form with a space - the replacement text is converted the same way. All spaces of a key are converted at once; text without %20 is never changed by the extra key. Keys you entered yourself in both forms are not doubled. The option is off by default; in a command-line run use <KEYSPACEVARIANTS>True</KEYSPACEVARIANTS>.
• Shorter option names on the Search & Replace form. Backslash to slash is now \ to /, Slash to backslash is now / to \ and Keys: space = %20 is now Keys: Space<->%20. The options work as before.
• Convert "\ to /" and "/ to \" only with links. The options are available only when an area with links is selected (now also Target Path on the LNK/URL page and SharePoint Summary Links). Before, an option that was switched on stayed active even after it was greyed out - for example with only Update text selected it converted backslashes in the whole text of the document. Now it is switched off when no area with links is selected.
• Replace Preview on local and network locations now opens documents only for reading (previously they were opened for writing). This helps a replacement preview when there are read-only files.
• The search&replace log no longer shows additional keys that came from the search string length and were not used for replacements but were misleading.
• Search & Replace without documents to process starts an Ultra fast scan. When you start Search & Replace without scanning first, ReplaceMagic runs an Ultra fast scan of the selected location and then the replacement. If ReplaceMagic had already marked the session as scanned (for example after the Search & Replace window was closed without a scan) the replacement started with an empty list and processed 0 documents. Now an empty list (no documents, lists or menus to process) always starts the Ultra fast scan first; the log says so.
• Much faster replacement with long lists of search/replace keys. ReplaceMagic now keeps the keys as a list in memory and first checks whether a key occurs in a link at all, before it prepares the replacement. With thousands of keys, a workbook with many links is processed in seconds instead of minutes, so far fewer documents reach the "Maximum time to process document" limit. The result of the replacement is exactly the same as before. Tip: an Excel workbook stores the address of a hyperlink encoded (a space as %20). If your keys contain spaces and should also change hyperlink addresses in Excel, switch on "Keys: space = %20".
  
  

Document types

• Documents are now saved to memory first. If saving fails, the original file stays unchanged, so files can no longer end up with 0 bytes. This applies to all supported document types.
• Improved processing of FieldCodes in Word documents.
• PDF Actions of type GoToRemoteAction are now processed (requirement: the linked file needs to exist on the new target location).
• Pivot table sources keep a "%" in the file or folder name. When the source of a pivot table was replaced and its path contained a "%" (for example Mob Data ex60% V2.xlsx), the new source was saved with a wrong character instead of "% V" and the pivot table could no longer find its data. The name is now saved exactly as it is - for links to SharePoint and for paths on a file server.
• Linked Excel charts in PowerPoint are now written with their real path (previously URL-encoded with %20), which was the cause of the "path too long" error on save.
• Improved processing of .htm and .html pages.
• Encoding improved for text-like files.
  
  

Command-line (silent) runs and encryption

• Command-line (silent) runs with several Search & Replace keys. Since version 2026.1.0.9 a silent run without SmartMapper read the search and replace texts from the XML file in the wrong order when there was more than one key: with two keys the second key was not applied, with three or more keys wrong pairs were applied (the replace text of one key with the search text of the next) and some documents failed with Collection index must be in the range 1 to the size of the collection. Runs with one key, SmartMapper runs and runs started from the ReplaceMagic window were not affected. If you use silent runs with several keys, check the documents processed with versions 2026.2.0.0 to 2026.2.4.4.
• Command-line (silent) runs use the search strings from the longest to the shortest. The key pairs from the XML file were used in the order of the file, so a short search string at the top of the list could change the beginning of a value that a longer search string was meant to replace as a whole. They are now sorted by the length of the search string, longest first, as in an interactive run; search strings of the same length keep the order of the XML file.
• Several XML files in one start. When ReplaceMagic was started with several XML files (ReplaceMagic.exe a.xml b.xml), the second file inherited every setting of the first one that it did not set itself - for example the SmartMapper run and its keys, the selected areas (Formulas ...), Backup, Match case, Results in database or the SharePoint data - and every file after the first one ran an Ultra fast scan. Now every XML file starts from the same initial settings. Runs with one XML file (for example from the Task Scheduler) were not affected.
• Command Line Parameters: easier run selection. The run list now shows the run ID first (e.g. 16 - SmartMapper - Scanning ...), so it stays visible for long run names. The list shows only runs that already have Search & Replace keys; a tooltip and the message shown when no such runs exist explain that keys are created in the SmartMapper wizard.
• Export to XML config file for App-Only with certificate authentication.
• "Encrypt connect data" now really encrypts the user name and password. When the XML file for command-line (silent) runs was created with Encrypt connect data, the user name and password could still be written as plain text (with Copy to clipboard, and with Save without copying first), although the file said they were encrypted. Now they are always encrypted, also for additional SharePoint sites. An empty password (for example with App-Only authentication) no longer causes an error. Please create again the XML files that were made with Encrypt connect data in earlier versions - they may contain the password as plain text. Old files still work.
• "Encrypt connect data" now also encrypts the App-Only certificate password. In the XML file for command-line (silent) runs the password of the App-Only certificate (main site and additional sites) was written as plain text, even with Encrypt connect data. It is now encrypted like the user name and password. XML files created with earlier versions still work; create them again so that the certificate password is no longer readable in the file.
• User names and passwords with special characters. Encrypt connect data (command-line XML) and the encrypted mail settings now encrypt user names and passwords with non-English characters (for example č, ž, ä, é, €) completely. Before, such values were cut off. Values with English letters, digits and the usual symbols are encrypted exactly as before, so existing configuration files keep working; if a saved user name or password contains such characters, save the settings or generate the XML file again.
  
  

Database

• Database version 2026.3.0.0. The new IdSourceRunID column (Resolved Links) is added automatically on SQL Server, MySQL and SQLite, and is included in new databases. The same applies to the new columns ReviewLink and BlockDownloadLink (scanned documents) and, on SQL Server, Filename (Unresolved Links; MySQL and SQLite databases already have it).
• Database version 2026.4.0.0. New column RuleOrigin (mapping rules) and the columns SmartMapperVersion and KeysCreated (runs - which SmartMapper version created the keys, and when). They are added automatically on SQL Server, MySQL and SQLite and are included in new databases. If an update step fails, it is now retried on the next start.
• Database update from older versions. Databases on version 2026.1 or older now receive all required update steps in one go. Previously the update could stop after one step, and SmartMapper then refused the database with the message "This database predates 2026.3.0.0". Update errors are now written to the log.
• SQLite database update: no more data loss. Updating an SQLite database from 2026.1 to 2026.2 could lose data: mapping rules, resolved and unresolved links, Search & Replace keys, the FullFolderPath column and database indexes. The same could happen when an update step ran again on an already updated database. All data and indexes are now preserved:
  • the update verifies the copied data;
  • if anything goes wrong, all changes are rolled back and the database stays as it was;
  • steps that were already applied are skipped.
• The database is now updated automatically before a run writes its results. Before a scan with SmartMapper, a command-line (silent) run with Results in database or a SmartMapper silent run, ReplaceMagic checks the version and the table structure of the database and updates it if needed - silently, without a message window (the status bar and the log show it). Before, the database was updated only when Use SmartMapper was clicked, in the SmartMapper wizard or in the export windows: a silent run with an XML file created before the upgrade, or an older existing database chosen in DB Config, wrote into the old structure and the results were not saved (table ScannedDocuments has no column named UniqueID). If the database cannot be updated (for example no right to change the table structure, or a read-only database file), the documents are still processed and the log explains why the results could not be saved.
• Results with an apostrophe were not saved in the database. When the result of a link check contained an apostrophe (for example The remote name could not be resolved: 'www.example.com'), or a sheet name or a user name did (for example O'Brien), that row was not saved and the log showed SQL logic error. These rows are now saved.
• Export Run: the database update starts right after the "Database update required!" message, and "Database updated!" is shown only when the update succeeded.
• DB Wizard (SQL Server). After Create database, Next no longer says that the database does not exist (before, you had to go Back and click Check connection again). If the database exists but the user has no access to it, Check connection now says so instead of trying to create it again. Create database shows the real reason when it fails (for example CREATE DATABASE permission denied), creating the tables again in an existing database works, and Drop database no longer fails with "currently in use". Long messages in the status bars and in the Create database list are shown in full in a tooltip or with a double-click.
  
  

Import and export of scan results

• Scan Documents export. Without the "Include" options the export had a fixed set of 30 columns and showed the UniqueID value in the spGUID column; with the "Include" options it produced only the "Scanned documents" sheet; the detailed export showed the owner in the "Last Editor" column. All fixed.
• Import of exported CSV files with commas in names fixed. The CSV export for Import files puts values that contain a comma in quotation marks (for example the folder "H,S&E Advisor" or a user name "Smith, John"), and so does Excel when such a file is edited and saved. Import files did not remove these quotation marks, so documents with a comma in the folder or file name could not be opened, and SharePoint metadata with a comma (for example the names of Created By / Modified By users or check-in comments) was read with the quotation marks. Now these values are read correctly.
• Import of scanned documents: values over several lines. A value in the CSV file that contains a line break (for example a comment in the SharePoint metadata file SP_...) split the document into two broken rows. Such values are now read correctly. If a quote in the file is never closed, the rows are read one by one as before.
• CSV import of older export files no longer writes the folder path into columns that did not exist in the file (FullFolderPath, UniqueID, spGUID).
  
  

Other

• Complete lists of skipped documents and replacement problems. When many documents were processed in parallel and several of them were skipped or had a problem at the same moment, some entries could be missing from the list of skipped documents or from the Search & Replace problems, and the same number could appear twice. The counters were correct, only the lists were incomplete. Now every skipped document and every problem is listed, numbered without gaps or repeats. The documents themselves were not affected.
• Parallel processing stays at the set number of documents. When ReplaceMagic stopped a document because it took too long (document processing timeout), or when you stopped it in the Active Documents window, the document was counted as finished twice. After several stopped documents ReplaceMagic started more documents at the same time than set in Parallel document processing - for example many large presentations at once - which slowed the computer down and could cause even more timeouts. Now a stopped document is counted once, and the number of documents processed at the same time (Used parallelism in the Active Documents window) stays exact. In the log, a document you stop yourself in the Active Documents window is now shown as Stopping document, manually stopped in Active Documents - before it looked like a timeout.
• Long paths: the log now shows whether long paths are supported and which setting is required.
• Updated license agreement released (see installation folder).
• Faster replacement with a long Include/Exclude list. When the filtering rules are set for replacements (Processing Filter: "Exclude/Include rules for: replacements"), ReplaceMagic now applies the list once, before the replacement starts. The number of documents at the start of the replacement, the progress bar and the Replacement dashboard now show only the documents that are really processed. Before, they showed all scanned documents. Long lists, for example an Include list with thousands of documents, no longer slow down the replacement. The log now has one summary line, for example "4,046 of 239,512 documents are processed, 235,466 skipped by the filter". It no longer has one line "Document skipped due to selected filters!" for every document. The filter settings (Include or Exclude, files / folders / files and folders, number of strings, scanning / replacements) are now written at the top of the log. Empty strings in the list are ignored. Before, an empty line in an imported list matched every document.
• SmartMapper: warning for documents with keys that are not found. Before a SmartMapper replacement, ReplaceMagic now checks whether all documents with keys of the run are in the documents to process. SmartMapper finds the keys of a document by its path, so the links in documents that were not migrated, deleted, moved or renamed by the migration tool (for example ShareGate replaces "#" and "%" with "_") were not changed, and nothing showed it. Now the message "Documents with keys not found" shows how many documents are missing, an example, the possible reasons and what to do. The replacement then continues with the other documents. The full list is saved next to the log. A command-line (silent) run writes the same information to the log and shows no message.
• Log of the replacement with a processing filter. At the end of the log, "Number of processed documents/lists" now shows the number of documents after the filter, the same number as in the Replacement dashboard. Before, it showed all scanned documents.
• Database export with long date formats. Saving a scan or a SmartMapper run to a SQL Server or MySQL database failed when the Windows date format is long, for example the US format with AM/PM ("String or binary data would be truncated ... column 'StartTime'"). The date and time columns are now longer, and existing databases are updated automatically before the next export.